Iab europe road_to_programmatic_white_paper_july_2015_15.07.15

Road to Programmatic
An IAB Europe White Paper
July 2015
europe

ROAD TO PROGRAMMATIC WHITE PAPER
1. Introduction and Executive Summary 2-3
2. Scenarios for Decision Making
2.1 Decision Tree
2.2 Operational Model
2.3 Toolkit Selection
2.4 Inventory and Sourcing Methods
2.5 Thinking Cross-Channel and Cross-Device in a
Programmatic World
4-6
7-16
17-29
29-33
33-37
3. Data 38-51
4. Further Strategic Considerations
4.1 Quality
4.2 Programmatic Mobile
4.3 Programmatic Video
4.4 Understanding the Value Chain
52-56
56-59
59-61
61-65
5. Current Adoption of Programmatic Advertising 66-70
6. Conclusions and Recommendations 71-72
7. White Paper Contributors 73-75
8. Glossary 76-78
Contents

2
Introduction
In the first six months of 2015, programmatic advertising has been at the
centre of discussions across the advertising industry. From Advertising Week
in London and Interact in Berlin to the Festival of Creativity in Cannes, it has
been never been far from the headlines. Indeed, at IAB Europe the small
Programmatic Trading Task Force formed in 2014 with a handful of members
is now a full Committee with a significant membership across the European
market.
Where in the first IAB Europe Programmatic White Paper the focus was on
defining programmatic and explaining why it was the subject of so much
discussion, this White Paper provides an analysis of the decisions facing
advertisers, agencies and publishers in considering how to capitalise on the
programmatic opportunity and address the accompanying challenges.
The Road to Programmatic is intended to be a reference for stakeholders
wanting to know more about programmatic strategy, offering guidance from
industry experts on the data, technology, operating models, and related issues
that affect and determine opportunity, risk and success. Here, IAB Europe
aims to deliver a comprehensive set of information that will be relevant for
different stakeholder groups and enable the empowerment and consensus
required for a profitable road to programmatic
Since the first white paper IAB Europe has delivered webinars, conference
presentations, workshops and a market sizing to aid to the development of
understanding of programmatic. IAB Europe would like to extend its thanks to
the Programmatic Trading Committee members whose energy and dedication
have made this possible.
Executive Summary
In 2015, the media and advertising business in Europe remains complex,
inefficient and often combative with tensions between advertisers, agencies
and publishers increasing rather than declining as the economy improves.
Although programmatic is often criticised for adding to that complexity and
frequently only talked about in terms of automation and cost reduction, this
paper demonstrates how it has the potential to overcome inefficiencies of
delivery, fragmentation of audience and lack of understanding to unlock
1 Introduction and Executive Summary

3
hidden value in the data held by advertisers and publishers alike. It shows
how programmatic can foster greater collaboration, thus moving the focus
from the cost of the media to the value of the audience.
The Road to Programmatic explores the operational models of agency and
agency trading desks, independents and independent trading desks,
programmatic services, client in-house and hybrid. It demonstrates the
different levels of expertise and involvement required of the different
stakeholders for each approach. This enables advertisers, agencies and
publishers to develop their strategies and maximise value.
The paper illustrates the importance of a fundamental understanding of the
target audience and how to use audience data with programmatic tools to
develop successful strategies, be they with either brand or performance
objectives. It reminds the reader that both companies and consumers benefit
when relevant advertising and content is delivered at the right time, in the right
place and to the right person. Measurement of success needs to include
attribution so that each element of the digital marketing mix is able to be
quantified and valued. Measurement also needs to be expressed in terms that
are relevant across all media channels and IAB Europe’s Blueprint on Online
Audience Measurement and Ad Effectiveness Metrics and KPIs is referenced
to this effect.
Ultimately the real value of programmatic is reached when it enables insight,
collaboration and results and serves as a strategic tool for the future.

4
The decision trees on the following pages are intended
as a guide for buy-side and sell-side stakeholders to
evaluate whether programmatic is suited to their
business and to decide which operational model is
appropriate.
2
2.1
Decision
Tree
Scenarios for Decision Making

Contact: Alison Fennah – fennah@iabeurope.eu / Marie-Clare Puffett – puffett@iabeurope.eu
BRANDING /
PERFORMANCE
OBJECTIVES
+ Pay for leads,
sales and actions
for performance
objectives
+ Pay for
actions related
to branding
objectives e.g.
brand awareness
- Vendor takes
risk but control
and any upside of
campaign
RE-TARGETING
+ Provider may
have specialised
tech to maximise
ROI
- Siloed
optimisation
on last step of
purchase process
AUDIENCE
BUYING
+ Vendor may
have large
proprietary
audience data
pool
- Quality of
Audience data
hard to validate
I am a digital first
business with lots of first
party data. How do I use
this to my advantage?
I am an established
global advertiser with
millions of consumers.
How do I increase their
engagement with my
brand?
I don’t have a large
advertising budget so
can’t afford blanket
coverage. How do I
reach the people most
likely to buy?
Our consumers are
moving to mobile. How
do we achieve scale
and consistency across
screens?
Effectiveness of TV,
search and social
has been reduced by
media fragmentation.
What’s next?
Decision-making Tree for a Buy-side Programmatic Trading Strategy
My business
doesn’t have
an integrated
agency support
or in-house digital
team’
My business
wants to focus
on one aspect
of demand
generation
My business
needs quick results
without needing to
understand how
they are delivered
PROGRAMMATIC PRODUCT
Buy defined products like in Ad Network model
+ Quick to start and easy to switch
- Limited transparency and control
AGENCY OR
AGENCY
TRADING DESK
+ Integration
across all
channels,
established service
model
- Understanding
agency fees and
media margin
INDEPENDENT
TRADING DESK
+ Specialised
agency sometimes
with own tech
- Siloed channel-
optimisers,
- Understanding
agency fees and
media margin
MANAGED
SERVICE BY
TECHNOLOGY
VENDORS
+ Know their
tools best
- No unbiased
view on
technology
opportunities
- Does Provider
earn on fee only
or also on media
margin?
My business
manages
integrated
operations across
channels without
easy access to first
party data
My business
values the speed
of results and the
understanding of
what works and
why
My business
is comfortable
negotiating
and managing
strategic partner
relationships
PROGRAMMATIC SERVICE
Outsource operations
+ Quick to start with good transparency and control
(depending on contract)
- Limited learning
Whichprogrammatictradingoperationalmodelisbestsuitedtoyourbusiness?
My business is a
growing ‘digital
first’ operation
from consumer
acquisition to sale
My business has
rich first-party data
and the resources
to analyse and
apply this to
marketing activity
My business has
the resources and
management
commitment to
build for the long-
term opportunity
My business is
able to attract and
retain the right
talent
BUILD TEAM
+ All know-how
integrated into
organisation
- Hard to find
talent, risk of siloed
view on market
SOURCE DSP
AND DMP
+ Good choice of
tools available
- Own team
needs to keep up
to speed with new
tools
BUILD
COMPETENCE
IN-HOUSE
+ Good
connection of
CRM, sales,
consumer service
- Significant
training effort
to bring outside
knowledge in
IN-HOUSE
Build team, licence or build technology
+ Highest transparency and control building internal core competence
- Slow to start with higher fixed costs and limited scope to change
Isyourbusinesssuitedtoa
programmatictradingstrategy?
Business is suitable for a programmatic trading strategy
europe

NETWORK PRODUCT
(Pass inventory to a third party network for them to
sell direct or programmatically
+ Relatively simple to orchestrate and low cost,
with minimal channel conflict
- Limited transparency and control
IN-HOUSE
Take control of the technology and strategy
+ Build a differentiated long term strategy that
minimises channel conflict
- Requires time and financial investment to get right
OUTSOURCE TO AN SSP
Engage an SSP to manage programmatic sales on your
behalf
+ Quickest way to begin true programmatic trading
+ Relatively simple to orchestrate with better transparency
and control than a network
- Potentially higher costs than the network model with
less transparency or
control than in-house operations
I regularly have
inventory that
remains unsold
I have buyers that
want to trade with me
programmatically
I want to use
audience data to
increase revenue
I want to improve
the efficiency of my
ad operations
Monetising on
mobile is
problematic
Decision-making Tree for a Sell-side Programmatic Trading StrategyWhichprogrammatictradingoperational
modelisbestsuitedtoyourbusiness?
Isyourbusinesssuitedtoa
programmatictradingstrategy?
Business is suitable for a programmatic trading strategy
europe
Contact: Alison Fennah – fennah@iabeurope.eu / Marie-Clare Puffett – puffett@iabeurope.eu

7
With a growth figure of 111% in 20131
programmatic
advertising has increased in importance over the past two
years and risen from being delegated by client-side
marketers to the specialist media traders working on their
advertising, to having a central role in every online brand
strategy.
The results of research carried out amongst global CMOs towards the end
of last year by AudienceScience in conjunction with the IAA and WARC2
,
showed that 46% of all advertisers and 56% of large advertisers (spend
£40m+) believe that a deeper in-house knowledge of digital advertising
technology will be the most important factor in bringing them digital
advertising success.
That transition, however, is not without its challenges. Programmatic is
inherently complex, as the digital advertising world is far more fragmented
than search or social. The move from buying Media to buying Audience is a
conceptual shift and the scale and speed of data and transactions involved
can be mind-boggling. Capitalising on the opportunity to engage a potential
consumer, before they declare intent and where and when it is relevant to do
so, requires advertisers to rethink the way they use digital advertising and for
most that requires a rethink of the operational model.
Traditional Media Buying
The ‘marketing lunch’ may be fading into history, but much of the way digital
media was bought and sold until recently owes more to this era than the
Internet. Campaigns are planned out months in advance, media is bought to
cover the expected audience and duration and the campaign is set up and
run with only limited options for ‘in-flight optimisation’. The operating model
has barely changed in decades; with the marketer briefing the agency,
whose media buyers then execute the plan. The agency value-add in this
world is clear – better media rates and a team of specialists to do the
campaign legwork.
In the old world, media is a commodity and digital advertising is a game of
scale. In a programmatic world, this changes. Audience becomes the focus,
and therefore the quality of data and insight around that audience is
paramount. For much of the performance market, financial scale has now
1 IAB Europe Programmatic Market Sizing Study (September, 2014), IAB Europe -
http://www.iabeurope.eu/news/iab-europe-values-european-programmatic-market-21bn-2013
2 The Shifting Digital Landscape White Paper, Audience Science -
http://demand.audiencescience.com/whitepaper/shiftingdigitallandscape
2.2
Operational
Model

8
given way to the breadth of data insight and the agility to use it effectively.
For branding campaigns, while publisher relationships and price negotiation
still matter, buyer success depends more on the ability to efficiently reach
and engage the target audience, than to reduce the cost of media.
Unfortunately, the tendency to treat media as a commodity and the role of
procurement in the advertiser / agency relationship means that we are going
into the programmatic era with an operational model focussed on driving
down cost without due regard to maximising the value of each impression.
As a result, many marketers are faced with a choice between a model they
understand, but don’t fully trust or venturing into new programmatic models
that they don’t yet fully understand.
Programmatic Operational Models
Before deciding on an appropriate operating model for programmatic, it is
important to consider a starting point. The ‘decision tree’ in section 2.1 is
intended as an outline guide to evaluating an appropriate ‘road to
programmatic’. If an advertiser is a digital-first business, where consumer
acquisition online is the core business process, they probably have built that
capability in-house. However, if it is a large scale traditional advertiser they
probably have a multitude of different approaches across brands and
countries supported by many different agencies and have probably spent the
better part of the last 20 years trying to marshal and manage data with
limited success. The aim of this paper is not to propose a ‘right’ answer, but
to outline the possible operational models and some of the questions to ask
as stakeholders embark on their journey or progress to the next stage.
The challenge in selecting the best fitting operational model for an advertiser
comes from the complexity of a fast-evolving landscape can be evaluated by
first answering two questions:
1. Who should be involved in delivering the programmatic solution?
In-house teams, a managed service by an agency or technology
provider or the provider of a packaged product?
2. Which business model of that provider is the right fit?
A simple and fairly risk-free product model or a more complex but fully
transparent service model?
Before looking at the breadth of partners that support advertisers in their
programmatic setup, lets first look at the second question of the business
models those partners can offer.

9
When buying media programmatically, there are essentially two business
models to choose:
a) Programmatic Products
b) Programmatic Services
Programmatic Products: pay for results but have less control
Programmatic products were the first offers on the market. The vendor
combines media inventory, technology, data and experts into defined
outcomes. Core examples are re-targeting networks or audience buying. The
buyer sends an insertion order, paying a cost per thousand/ mille (CPM),
cost per click (CPC) or cost per acquisition (CPA) for a defined outcome,
whether that is bringing back users to the advertiser’s site that abandoned
the shopping process (re-targeting), reaching users of a defined audience
profile i.e. men aged 20-59 in-market for a car (audience buying) or
delivering leads (cost per acquisition buying).
The vendor delivers that outcome, buying untargeted impressions on CPM
and converting them into the desired outcome. If the vendor is successful,
they pay less for the raw media than the advertiser pays to them for the final
product: the vendor keeps that margin. But the vendor also carries a risk.
It is typical for the product model, that the vendor does not disclose all
components of the media buy. They take the risk of converting impressions
into outcomes and have as much control as possible on which media to buy,
which data to apply, which algorithm to use and in the case of retargeting
also often, which creative to use.
Some people call this model ‘arbitrage’. However unlike the original concept
of arbitrage in financial markets, the programmatic product vendors do not
only buy and re-sell impressions but add data and technology and often take
a conversion risk.
The core benefit of programmatic products is the ease of buying them, the
low risk and often very good results they provide. There are no upfront costs
and often also no commitment, so the advertiser can shift budgets as the
products perform.
The core disadvantage that some advertisers see is the lack of control. The
vendor does usually not disclose which media they bought for which price
and how they optimise. Some vendors may allow the advertiser to deploy ad
verification tools in the buy to control brand safety, but as in any typical
product, the buyer has limited control about how the product is produced.

10
Programmatic Services: pay for all components of the programmatic
setup and get full control
For advertisers who want to take full control, the service model may be a
better fit. Here the advertiser pays separately for the media inventory bought,
the technology, third party data and the team of experts bringing it all
together. This model is like the traditional agency model, where media costs
are a transparent pass-through that can be audited and a fee is paid for the
service. Now advertising technology (demand side platforms - DSPs,
verification tools, potentially a data management platform - DMP) and third
party data costs become significant additional cost components, but in a true
disclosed service model, those costs are passed through transparently as
well.
Despite a lot of discussions about automation in programmatic advertising, it
can take significant time and effort to manage programmatic campaigns, as
the choice of options on inventory, data and optimisation strategies is much
higher and the vast amount of data require both technology and humans.
The true programmatic service model gives the marketer control and
accountability on the technology choice as well as access to all of the
reporting and performance data. The service provider then delivers the
expertise to get the most from the technology. There have been several
high-profile examples of major consumer businesses adopting this model
where their agency provides the staffing and services support, which is why
the World Federation of Advertisers (WFA) has at times referred to this as
the brand trading desk – a dedicated agency resource wrapped around the
clients chosen technology.
The advantage of the disclosed programmatic service model is obvious; the
advertiser has full transparency and control. However, there are also some
disadvantages. With ownership comes full accountability for all of the fixed
costs in the process.
Advertisers have learned in search that it takes more man-power to spend a
Euro than in traditional display, even though search is only one to two
vendors, few advertising formats, few targeting options and brand safe
placements. But the volume of data and the ability to micro-manage
campaigns means that optimisers have to be always-on, otherwise they
waste budget. In programmatic there are many more levers to pull, and as in
search it does not work for most tactics just to put the bid-manager into auto-
control.

11
A consequence of taking control is that the advertiser is involved or at least
needs to understand which technologies are used (DSP, DMP, verification
tools), which media buying strategies are deployed (open exchange, private
marketplace deals) and how the campaigns are executed.
Another consequence of taking full control is that the advertiser does not
benefit from the audience pools that programmatic product companies or
agency trading desks have built. The advertiser also carries both the risk and
reward of optimisation: media and third party data is mostly paid for on CPM,
tech licence fees on CPM or percentage of media spend and the
employment of specialist teams. Whether that combination delivers the
desired results is the advertiser’s risk – but the advertiser also benefits from
all the optimisation outcomes.
When it comes down to who does the work, there are five current operating
models in the market:
 Agency
 Independent
 Programmatic Services
 Client in-house
 Hybrid
These models are explored in turn through the rest of this section
The Agency and Agency Trading Desk Model
If an advertiser is already using an agency to handle its digital advertising,
it’s a logical place to start the programmatic journey. However, not all
agencies were quick to see the changes that programmatic would bring and
there are considerable differences in levels of understanding and operational
scale. The early adopters within the client community found in the early days
of programmatic that they sometimes knew more about the technology and
its implications than the agency team working on their account. However,
over the last three years the agency world has moved quickly to keep up
with changes and then move ahead of them. As a result, the majority of
agencies will now offer a robust programmatic model and levels of
knowledge are far higher than they once were.
Within the major agency groups, programmatic capabilities were first built
inside agency trading desks (ATDs). These allowed the agencies to invest at

12
scale to achieve the necessary operational agility and apply what they
learned as the market developed for the benefit of all clients. These are now
highly evolved programmatic businesses, delivering campaigns across
industries and countries. As technology and skills reach critical scale,
programmatic expertise is flowing back into their related media and creative
agencies and the account management teams that most marketers will deal
with. As a result, the agency model remains the fastest and easiest route to
programmatic for many businesses. With access to a multiplicity of trading
partners, an agency trading desk provides a managed service that will
control the whole process.
Agency trading desks are owned by the world’s largest marketing services
companies, buying and selling online advertising space on behalf of a large
number of advertisers. Outside the real-time auctions that form part of
programmatic, they continue to buy media space in bulk, and therefore are
able to negotiate lower prices from media owners bringing together the best
of the traditional and programmatic approach.
The agency trading desk model is also good for advertisers that do not have
a large amount of, or need for, first party data; as the agency trading desks
have built their own data models and have access to extensive third party
data resources. Advertisers with significant, proprietary data have expressed
concerns over whether that data is used exclusively for their own campaigns
or is fuelling audience segments that are shared across agency clients. The
issue is less debated than it once was, but in selecting an agency partner it
is important to understand and make clear in the contract all issues relating
to data ownership and usage.
The transparency concern has also been raised by organisations such as
the World Federation of Advertisers (WFA), where it is unclear to marketers
working with some agency trade desks how the campaign is being managed,
where the budget is being invested and how much of the total spend is on
‘effective media’ vs. technology and services. Agencies have responded to
this in a number of ways and so it is important for the marketer to find a
model that is appropriate for them; reflecting their level of understanding,
data and technical capability.
As programmatic moves from being a line-item on the plan, to the majority of
the plan the advertiser’s understanding of the agency model becomes more
important. It is relatively easy to run a performance based campaign based
on a black-box product approach to media buying, as the results will speak
for themselves. It is far harder to deliver brand activity in this way, as it will

13
rely on more complex key performance indicators (KPIs), audience insight,
measures of viewability and positive context.
Agencies now operate non-disclosed product models, fully transparent
service models and hybrid models, so advertisers have choice to find the
most appropriate model.
Independents and Independent Trading Desks
As with every past marketing innovation, there were some in the industry
that saw the potential for the new technology before everyone else. In the
programmatic world, these are independent specialist agencies and trading
desks. Many started life offering programmatic expertise to both client-side
marketers and agencies as the new technology first emerged. They operate
in a similar way to the agency trade desks, offering advertisers a managed
service that can take control of the whole process, but can be more
experimental with the technology and strategies they use as they operate as
a smaller, more agile business. On the other hand they may lack the global
footprint and broad cross-market and cross-client learning that others may
offer.
The main difference is that independent trading desk ownership does not lie
within a media/marketing agency so there is no possibility of conflict of
interest within the wider agency group. It is unlikely an independent trading
desk will have the same kind of global media trading agreements to fulfil, or
be working with such a wide range of potentially competing clients. However,
because of their independence, they are much less likely to have access to
large, global deals with media owners and so the cost of media may at times
be higher. They are also specialists in one part of the digital landscape and it
may be harder to integrate programmatic activities running through an
independent with the wider marketing mix. However, there are some
excellent examples of independents delivering ground-breaking innovation
and doing a particularly good job of leveraging the clients other data and
insight to deliver results.
Many independents have invested in building their own technology rather
than licensing a third party development, which can give them more control
and flexibility that can be passed onto clients. At the same time, this can
make them slow to adapt when technologies improve around them and they
cannot invest to keep pace.
Because they tend to be smaller businesses than the major agencies, the
client will often find themselves in a stronger position to demand greater

14
transparency across the operation and have a custom solution for their
business.
Programmatic Service Model
The programmatic service model uses a different approach to the agency
model, in that an advertiser sources the different components and suppliers
required to carry out programmatic buying independently. They contract
directly with the technology vendors and use a range of service providers for
campaign delivery. This creates operational overheads within the client
business that must be weighed against a higher level of choice and control;
the ability to evaluate different options based on their own needs and
preferences; and access to some of the best of breed technology solutions
and partners.
However, using a number of different suppliers can mean that there is a lack
of integration across the end-to-end programmatic solution. When things go
wrong it can be difficult to identify the cause and difficult to resolve.
Although a popular concept, the programmatic service model has been slow
to take shape. This is primarily because of the lack of skills and expertise
across the market. Technology may sit at the heart of programmatic, but it
requires knowledgeable users to get the most from the technology. The
great majority of the available expertise resides either in the agency world or
with the technology companies.
The weakness for the programmatic services model is really the availability
and usage of data, through a fully integrated data management platform. A
fragmented programmatic model may be best for transparency and control,
but it leaves a number of unanswered questions around how to capture,
manage and apply the data from each stage of the campaign process. It
addresses many of the unique needs to a large and active advertiser but it
does not go all the way.
Client In-house
The term ‘bringing programmatic in-house’ is one that has recently been
widely used and can refer to different degrees of advertiser operating digital
ad buying internally. It can sound as though advertisers are cutting all
suppliers, partners, and agencies and managing the whole process,
including technology and media buying themselves. However, it is typically
not this clear cut.

15
The reason the in-house option has sprinted up the agenda for global
advertisers is partly because of concerns with current models, and partly
because there are now alternative options making it more feasible. For
example, an Enterprise model allow advertisers to run their entire digital
advertising programme using one platform which they pay a technology fee
for, with support from the system’s client service team – much like
advertisers similarly use software as a service (SAAS) solutions such as
CRM, web analytics and email service providers. All of these SAAS services
are contracted by advertisers, but access can be provided to agencies as
required to meet the client’s strategy.
The ‘client in-house’ operating model option is best suited to advertisers that
have the significant global digital spend and for whom consumer acquisition
or direct to consumer sales are core. That makes the up-front cost of people
and technology a long-term investment in sustained competitive advantage.
That advantage is derived from the impact that programmatic has on the
advertising itself, as well as the visibility, control and flexibility that the
advertiser then has to adjust and adapt its approach to meet market
opportunities. It’s an advantage that is built on large amount of real-time
data, and so lends itself to the digital-first, data rich businesses that are
perhaps already running search and social in-house. Indeed the synergies
with these operations can be significant.
The challenges of in-house are not to be underestimated though.
Technology is not standing still and an in-house team does not have the
same opportunities to test out new tools and working models as an agency.
Indeed, the drive to an integrated ‘marketing cloud’ seen in many in-house
functions could result in less technology choice than other models as a
‘group standard’ is chosen for all marketing technology over specific
advertising and programmatic needs. The isolated nature of the clients’
campaign operations may also limit their ability to attract, retain and develop
the best people and continue to learn from others as they would if working
inside the agency ecosystem.
Hybrid Model
The options described above can also be merged in a hybrid model. Many
programmatic solutions reflect more than one category – for example an
advertiser can have an in-house real-time bidding (RTB) team, who manage
an independent or agency trade desk. This means they don’t ‘hand over the
keys’ entirely, and maintain control while retaining flexibility.

16
Programmatic is developing quickly, and while the agency model remains
the preferred approach for many, the hybrid model is actually likely to be the
most dynamic over the next few years. There is no ‘one size fits all’ and so
the industry is likely to see continued experimentation with programmatic
models, as clients explore new working relationships and technologies.
Although ‘transparency’ has featured heavily in discussions through 2014,
the increasing use of hybrid models will move the industry forward in 2015.
Marketers are likely to continue to experiment in specific areas, perhaps
around performance, or retargeting while continuing to gradually evolve the
way they build their brands programmatically. In times like these, it is not
transparency per se that matters, but the ability of the parties involved to
have access to the information and to understand how to apply it. Essential
in that process is asking the right questions for the business.
Choosing the Right Solution
No one likes being forced to make a choice when they don’t feel that they
have enough information to make a good decision. This is particularly true in
programmatic, where the evidence of the compelling benefits of the
approach are mingled with headlines around brand safety, viewability and
fraud that can make it seem like direct control is the only safe and effective
way to proceed.
The best approach to defining the appropriate operating model is to start
with smart questions. Here are five suggested starting points:
1. Do I truly understand where my business is starting its programmatic
journey – our data, capabilities, reasons for change and desired
outcomes?
2. Can the businesses I am thinking of working with explain how they
make money – are they a technology or media company, or some mix
of the two?
3. Is my organisation happy to ‘pay by results’ or do we need a granular
understanding of where the money is spent?
4. Do I have the budget, people and resources to make the change that I
think is ideal, or can I get it if the business case is strong enough?
5. How will I measure progress and how do we know when we have
arrived?

17
2.3.1 Buy-Side Toolkit Selection
Here, two key tools available to buyers are explored;
demand side platforms and data management platforms,
the former focused on accessing inventory and the latter on
managing data to inform the buying strategy.
What is a Demand Side Platform (DSP)?
A demand side platform (DSP), is technology used to purchase advertising
in an automated way. It allows advertisers and agencies to manage online
media campaigns by facilitating the buying of auction-based media,
including; display, mobile, video, social, and native as well as audience data
across multiple inventory and data suppliers in a centralised management
platform.
As the market has evolved, so has the DSP ecosystem. Many DSPs now
categorise themselves as ‘Trading Solutions’ or ‘Marketing Operating
Systems’ as their function is broader than the title suggests. DSPs are not
only used to purchase advertising inventory but serve, track, and optimise
campaigns, analysing the data generated to provide detailed insights. They
aim to make the ad buying process more cost-effective and efficient by
removing many of the manual parts of the process such as negotiating rates
and issuing insertion orders. They work by matching audience data with
inventory from multiple publishers to serve ads that target consumers at the
optimum time and with the relevant creative. The technology uses targeting
and optimisation algorithms to establish the value of any given impression,
and automatically places appropriate bids to minimise the cost of each ad
placement. In addition they can employ machine-learning capabilities to
optimise ad spend over time, based on the best converting inventory,
context, and audience segments for the advertiser.
How Does a DSP Work?
Publishers make their ad impressions available through marketplaces called
ad exchanges. An ad exchange is a big pool of ad impressions, and
publishers release their ad impressions into the pool making them available
for purchase. DSPs then automatically decide which of those impressions it
makes the most sense for an advertiser to buy based on information such as
their location, time of day, device type, ad position, and previous browsing
behaviour. Private exchanges are used by publishers to more carefully
control who can buy their inventory, and at what price. Instead of throwing its
2.3 Toolkit
Selection

18
ad impressions out into an ‘open’ exchange and letting anyone buy them, a
publisher might instead wish to offer them to a handful of pre-approved
advertisers, agencies or suppliers. They are also referred to as Private
Market Places (PMP). The price of impressions is often determined by a
real-time auction, through a process known as real-time bidding. That
process takes place in milliseconds, as a user’s computer loads a webpage.
Are All DSPs the Same?
No, not all DSPs are the same. Even the definition of a DSP is continually
changing in response to the needs of the marketplace, and there is a
continued move from vendors to the vision of a single platform approach to
programmatic buying. Popular analyst reports evaluate each platform
against almost 50 different criteria.
In addition to the basic concept of automated ad buying, bringing efficiency
and scale, technology is increasingly able to power cross-channel strategy,
remove silos between channels, and optimise every interaction across all
addressable media.
One DSP or a Multiple DSP Strategy?
Some companies choose to partner with just one DSP but some will use
more than one specialist DSP for the same campaign. The single DSP
approach offers a shorter learning-curve and real operational efficiencies,
but may not offer the full functionality of working with several specialists,
where each is designed to do one thing, such as mobile or video buying,
extremely well. Where a multiple DSP approach is used, care must be taken
not to create competition between them in bidding for media, or the
campaign will drive its own inflation as the algorithms bid against each other
for the same audience.
There are generalist and specialist DSPs within the ecosystem. Generalists
operate across all channels (display, mobile, and video), whereas specialists
tend to focus on one channel, for example mobile or video. Both have strong
market propositions.
Programmatic mobile has grown extremely fast over the last two years (see
section 5.2) and has spawned specialist mobile DSPs. These technologies
are built purely with mobile in mind and often have more comprehensive
access to inventory and reporting capabilities that address the challenges of
tracking media performance on mobile devices.

19
Video DSP technologies offer precise, sophisticated, brand safe targeting
alongside guaranteed video views, audience and site selection and even
data-driven segmenting. Inventory is a key differentiator among DSPs in the
video sector.
Key Considerations for Choosing a DSP
Some key considerations for choosing a DSP are outlined below.
Firstly, business focus and strategic alignment:
 How does the DSP make money? In a crowded ad tech market there
are many business models available. Some DSPs are purely a
technology provider that will provide training and support but no
manpower to deliver campaigns. Others offer a fully managed service,
built on their own tech while others lead with manpower operating
across a range of preferred technology partnerships. In every model, it
is also important to understand if the DSP is making a margin on
media to fund these services, or if they are independently costed. A
pure technology company typically operates on gross margins of 15%
or less. Media businesses will show margins of 40-50%.
 What is their customer service like? Understand who will be
working on the account, where their offices are located, and what
support will be needed for the size of the team. Local market support
is usually very important especially if the level of experience in-house
is moderately low. It’s worth checking what the support costs are and
how they are tiered as some DSPs charge for support beyond a basic
level. The company should be confident that the chosen provider has
the right cultural fit with their company and will invest time in making
the relationship work.
Secondly, technology capabilities:
 What technologies does the company already use, e.g. an ad
server, data management platform (DMP), dynamic creative
optimisation (DCO), tag management solution, attribution
provider, brand safety, CV tools etc.? If there are technologies and
providers already used as part of marketing activity then the company
should not change everything at once. The company should consider
what technologies they already work with and plan on working with and
understand the integrations in place with each of the DSPs being
considered. The biggest problems that occur when rolling out new

20
platforms is finding out that another technology provider cannot work in
a sustainable or manageable way with current partners. This should
not be underestimated and should be explored extensively when
selecting a DSP, particularly if a number of partnerships are already in
place.
 What unique functionality does each DSP have that is relevant to
the advertiser campaign requirements? What is the product
strategy moving forward? These are important questions to ask.
There will be varied answers to these questions so consider if the
answer is actually important to the objectives. Understanding the
product roadmap is something that should be weighed up when
entering into a DSP partnership as it will provide visibility on what may
be available in the short to medium term.
 Another area of difference between DSPs is the data they have
access to and more specifically the level of insights and
reporting. The use of data to provide high quality reporting and
insights is dependent on many factors so understanding what is
available, what is important , and how it is generated is key. The
importance of data is well documented and there are a few things that
need to be considered. The question to ask is; what access is
available and what is needed? This can range from wanting access to
very granular log level data, the ability to feed into other technologies
or accessing them through APIs or through an interface. It is important
to understand who ultimately owns the data and how is it used by the
DSP. Does it benefit others or does the advertiser retain complete
ownership?
Plus, specific operational factors:
 What are the campaign KPIs? Are they direct response or
branding? What is the primary objective for the campaign? It may
sound obvious, but when an advertiser is starting the process of
deciding which DSP or DSPs they intend to work with, understanding
the clear goals of the campaign will inform their decision making.
There are performance metrics that will be common with all DSPs, but
some will be stronger at delivering against specific goals. This is
something that is very hard to determine before starting to use a DSP
and is one of the reasons advertisers start looking to use more than
one provider. There are other metrics that an advertiser may want to
measure activity against for example; completion rates, gross rating
points (GRPs) or viewability – and some DSPs may be stronger than
others in each of these areas.

21
 Which markets is the company running in? Different DSPs have
integrations with different inventory sources (supply side platforms /
exchanges) and data providers. If there are local market providers that
are important to the campaign objectives they should be considered.
There is a lot of inventory and data available to DSPs but there are
local differences when looking at less common markets.
 What is the budget? Obviously, when thinking about planning
marketing activity the advertisers needs to have an idea of budget.
This will help to evaluate the commercial side of any DSP arrangement
and decide whether it makes sense to use a managed service or a
self-service model. DSPs have different recommended spend volumes
in different markets to ensure they achieve the best performance from
their proprietary algorithms so this is also something that should be
considered.
 The mix of media an advertiser is planning to run may affect what
they require from a DSP, e.g. mobile specific, video, display,
social or native? DSPs started with a focus on display advertising but
as programmatic has grown in both mobile and video, naturally there
are DSPs that have been developed specifically for these channels or
have evolved from companies that specialised in these areas before
programmatic took off.
Lastly, brand safety is a key topic to address. Brand safety measures are
tackled differently by each individual DSP. Some DSPs build internal safe
guards into their technology. Others integrate third party brand safety
providers alongside their own vetting of inventory providers.
The strict vetting procedures DSPs have in place to protect brands can be
used in conjunction with blacklists, whitelists, and third party brand safety
suppliers. Always ask whether the DSP has been accredited by industry
trade bodies and look for their seal of approval.
IAB Europe, the World Federation of Advertisers (WFA) and the European
Association of Communications Agencies (EACA) have a work stream
focused on quality topics, with the aim of improving the digital advertising
environment. Brand safety, non-human traffic, viewable ad impressions and
reach of intended target groups are a key focus alongside ad formats, site
environments and creativity. The European Digital Adverting Alliance
(EDAA) and the UK specific Digital Trading Standards Group (DTSG) seals

22
are other examples of initiatives to create best practice in brand safety.
Accredited companies agree to open up their processes and undertake
annual audits with independent auditors. There are different levels of EDAA
certification, speak to the trade bodies about the audit process and
understand what exactly it is that the seal is endorsing.
It is important to have visibility to be able to make informed decisions.
Understanding how inventory is being screened provides the opportunity to
exercise a level of control over inventory sources and what is and is not
acceptable. Blanket vetting procedures across all suppliers may mean losing
out on valuable inventory pools from suppliers who already have strict
measures in place and meet the highest industry standards. The risk here is
losing out on valuable unique inventory that can impact on audience
penetration and performance.
Summary
Rather than looking specifically for a DSP, which is only one part of a more
sophisticated puzzle, it can be helpful to think about it in the context of the
wider communications strategy and allow the strategy to drive technology
choices. However, there are some broad themes that it is useful to address
with shortlisted vendors:
 A long-term solution: technology evolves rapidly. As such, a
technology partner needs to provide a solution that allows for
customisability, modularity, and flexibility. This enables users to build
solutions tailored to their unique business needs.
 Data control and activation: the technology of choice should allow
open integration of data management and media execution
capabilities. It should utilise first-party data such as CRM, point-of-sale,
and social data; combine it with second and third party data, and
activate that data in real-time to produce targeted and effective digital
marketing campaigns.
 Support: implementation of programmatic technology is not a quick or
simple process. Client service and ease of use should be considered
alongside the capabilities of the technology. There is no point having
rich functionality if the team in place cannot capitalise upon it.
 Education: a DSP provider should provide users with educational
resources to help them understand the array of technological solutions

23
on offer and provide much-needed clarity in the programmatic
ecosystem.
 Collaboration: the ad tech ecosystem is complex, so a DSP should
collaborate with other companies in the marketing supply chain,
allowing their clients to efficiently execute their digital media plans in
ways that balance the powerful tools and potential with simplicity and
ease of use. It should connect users to the best-of-breed partners and
technologies within the media, technology, and data industries.
What is a Data Management Platform (DMP)?
A data management platform (DMP) can combine online behavioural data,
socio-demographic data with offline data to create a single, centralised hub of
consumer intelligence. With all offline and online data in one place, questions
such as the following can be asked: who are the company’s most valuable
consumers, and how do they behave in both the digital and physical worlds?
With a DMP, instant access to actionable feedback provides a solid basis for
testing new strategies and new tactics. The data may paint a picture that
confirms intuition; empowering marketers to make further refinements. Or the
data may be surprising, leading marketers to take a different course of action.
Either way, they know that their insights are based on facts, resulting in new
marketing confidence and more decisive action.
Working together: the DMP informs the decisions that are implemented
through the demand side platform (DSP), in the context of a full data strategy.
Because data gathering and insight is increasingly a collaboration between
buyers and sellers, this whitepaper deals with data in more depth in section 4,
but the DMP is noted here as an essential part of the data driven marketer’s
toolkit.
2.3.2 Sell-Side Technology Selection
What is a Supply Side Platform (SSP)?
A supply side platform or sell-side platform (SSP), is a programmatic
advertising technology platform for publishers, which enables them to
optimise the monetisation of their inventory and maximise yield, whilst
protecting their reader experience and consumer data.
In order to sell their websites’ advertising inventory programmatically,
publishers require a technological interface to establish the connection to the

24
auction. In the exact same way, stock exchanges in various countries are
powered by technological platforms, such as OMX.
An SSP is connected to multiple demand sources, which include DSPs, ad
exchanges and advertising networks. However, as the acronym suggests,
SSPs service the supply side – the publishers.
The ‘Traditional’ SSP
Historically, the SSP has served to manage and optimise secondary revenue
streams, monetising inventory a publisher does not sell through (usually)
premium direct (‘IO-based’) sales activity. It is a platform that serves to
provide publishers:
 turnkey connection to a broad mix of buyers, e.g. programmatic
demand sources such as DSPs, ad networks, and trading desks;
 optimisation capabilities that ensures the highest price is attained from
each impression monetised; and
 is integrated with internal business management systems (e.g. the
primary ad server)
In addition to connection with external demand sources, other ‘non-premium’
publishers’-specific sales might be incorporated in the SSPs optimisation, as
might be the case when a publisher is writing ‘direct’ performance deals with
cost per click (CPC), cost per action or cost per acquisition (CPA)
advertisers. Further, in recent years, data has become an important element
to almost all digital marketing, as such partner integrations have been
required, integrating with publisher or buyer data management platforms
(DMPs), retargeting firms. Beyond that, there are myriad other connections,
such as connectivity with rich media vendors.
Looking Forward
However, as the size of the European programmatic trading market
increases (€2.1bn in 2013)3
the role of the SSP is beginning to change. This
growth of programmatic is changing publishers’ sales strategies. Similarly, it
is changing the roles of the various platforms they have relied on for
monetisation.
So, ‘table stakes’ for an SSP, historically, has included: buy-side
connections; campaign and revenue optimisation and related analytics to
3 IAB Europe Programmatic Market Sizing Study (September, 2014):
http://www.iabeurope.eu/news/iab-europe-values-european-programmatic-market-21bn-2013

25
monitor and manage performance; and connectivity into related third-party
systems and service providers. In itself, that is a complicated footprint.
However, looking forward this will become more complex. As publishers
consider an SSP for today’s market there are several additional important
considerations. These include business focus and priorities; account
services team and approach; platform scope and vision and strategic
alignment with the publisher when programmatic becomes the core of the
business rather than an add-on.
Publishers no longer just have a growing number of impressions to manage,
they now need to factor in the different formats of these impressions and the
multiple screens where these impressions are delivered. SSPs help
publishers scale in this increasingly complex environment.
Private marketplaces (PMP) also known as private exchanges allow
publishers to establish a closed environment where they can nurture
preferred advertiser relationships by offering those advertisers first look at
certain inventory under previously negotiated terms. PMPs have grown in
popularity in Europe during the last three years and are many publishers
preferred method of selling their inventory programmatically.
Some SSPs are unveiling automated guaranteed functionality. Automated
guaranteed deals are for directly sold or “guaranteed” inventory and
therefore distinguished from traditional programmatic real-time bidding or
private marketplaces that involve bidding on individual impressions. While
automated guaranteed is still in its infancy, several industry projections
anticipate this market to grow quickly in the coming years.
Are All SSPs the Same?
There are two categories to consider:
1. Technology Oriented SSPs vs. Service Oriented SSPs
Technology orientated SSPs are based on a software as a service (SAAS)
model. They provide publishers with a platform, enabling them to manage
their inventory within a private exchange and according to a bespoke set of
business rules. Publishers are able to set the rules of the private exchange,
from the most basic – selecting and prioritising ad buyers, adjusting prices,
and using deal IDs (a universal identifier using numeric keycodes referring to
pre-negotiated pricing terms) – to the more advanced, such as exclusivity of
an advertiser in an advertising funnel. Technology SSPs allow publishers to

26
maximise their revenue while maintaining complete control over prices within
a brand safe advertising environment.
Service orientated SSPs manage the inventory selling process and
optimisation on behalf of publishers. They provide a generic programmatic
selling system in which publishers’ inventory is monetised according to pre-
defined rules, often with little customisation available.
2. SSPs using Server-to-Server Integration and SSPs using Waterfall
Process
Server-to-server SSPs auction an impression to every potential demand
source simultaneously in real-time. The SSP then identifies the best offer
made and automatically serves the appropriate advert. This process reduces
the risk of inventory wastage and the associated revenue loss, gaining the
maximum price for each impression.
Waterfall SSPs differentiate demand by waterfalling or daisy-chaining the
auction process. They first send an impression to a few ad buyers who are
likely to pay higher prices and then, if the floor price isn’t met, redirect the
impression down to the next level of buyers with lower floor budgets. This
process is repeated until the impression is finally sold, or goes to waste. This
process prevailed before open real-time bidding (RTB) standards were
drafted by the IAB, and is still used by some SSPs.
Apart from these two main distinctions, SSPs can also be differentiated
according to various criteria such as their media offerings (display, mobile
and video), the ad formats they offer, or their multi-device capabilities.
One SSP or a Multi-SSP Strategy?
As far as premium inventory is concerned, publishers typically choose a
single technology-oriented SSP in order to capture the highest value from
every impression. Using a single SSP that simultaneously sends a bid
request to all relevant ad buyers leads to increased auction competition and
monetisation efficiency.
When a publisher chooses to work with multiple SSPs, it can result in ad
buyers seeing the same impression offered several times, at different prices.
This has a negative impact both on the publisher’s brand and on its yield
optimisation.

27
Nevertheless, it is still common practice for publishers to monetise their
inventory by media category, making use of distinct SSPs for each of these
categories. For example, they may use one SSP for display and another
SSP for video inventory.
Key Considerations for Choosing an SSP
When selecting an SSP, publishers have to keep in mind their core interests:
Control: An SSP should enable publishers to take full control of the
programmatic process. Leveraging a full suite of customised tools,
publishers should be able to have control over ad buyers, inventory flow and
metrics such as time of day, landing page, position on the page, and ad
format. Ultimately, an SSP should protect a publisher’s main assets: its
users and its brand.
Transparency: An SSP should provide publishers with full transparency into
the auctions, buyers, bid density, and pricing on all media buys. Turning this
transparency into insights will then help publishers develop new strategies
according to market demand signals.
Value creation: An SSP’s technology should be agile enough for publishers
to translate their traditional media kit into programmatic advertising. The
higher the granularity of business rules that a publisher can implement in its
programmatic strategy, the higher the valuation of its inventory.
Sales channel management: An SSP will increasingly be a hub where
publishers can manage various monetisation channels, running direct buys,
where they keep an existing relationship with buyers, alongside indirect, while
maintaining a consistent yield.
Business focus: As programmatic advertising has grown, some SSPs have
evolved and will continue to do so. Publishers are well advised to see
strategic alignment in their SSP, where the technology and services is
focussed on helping them achieve their business goals and not simply the
revenue goals for the SSP. At the end of the day, an SSP should be, by
definition, sell-side biased, but those at the forefront are expanding the
definition to focus not just on seeking to effectively compete with buyers to
‘win’ but to collaborate with them so that both can win.
Account services: Further, while programmatic is an automated approach
to advertising sales, the services component of any SSP is as important as
the technology they provide to their clients. As the yield challenge becomes

28
more complex, the SSP can become a trusted partner that helps bring
demand to the publisher as well as fulfilling it.
Platform scope and vision: Perhaps most importantly, as programmatic
advertising has become more important to a publisher’s bottom line, it is no
longer really a ‘secondary’ or remnant channel. Real-time bidding (RTB)
exchanges gave way to private exchanges. With that came programmatic
direct and so-called ‘programmatic premium’ – effectively standard direct
sales business executed via the more efficient programmatic ‘pipes.’ This
translates to a programmatic business that is as important and potentially
more valuable delivering higher margins than a publisher’s direct sales. In
short, programmatic is premium, and it must be managed as such.
With this dynamic, no longer can the SSP sit in the second chair behind the
primary ad server. When evaluating SSPs, a publisher is well advised to
consider the provider’s current capabilities, development roadmap, and
ultimate platform vision. Any SSP today can be in a position to manage
direct sales in a manner consistent with ‘traditional’ ad servers, including the
requisite campaign management capabilities and inventory management
suite. It could also take pricing capabilities beyond simple bid-floor
management, dynamically creating/enforcing price barriers and discount
rules to govern the many competing sales models the publisher has going.
More importantly, the most advanced will have an optimisation approach that
can tackle the sales ‘channel management’ challenge – balancing open
market RTB, private exchanges, programmatic direct, negotiated business,
standalone cost per click (CPC) and cost per acquisition (CPA) deals, and so
on.
In this world, a publisher needs to evaluate each impression on its own
merit, against all possible ways to monetise it, considering inventory
constraints and campaign specific delivery requirements, and make the right
supply/demand matching decision in real-time.
Beyond this, the ‘right’ SSP should also have an eye towards the future.
Even the most knowledgeable of digital media players still has little visibility
into the truly long term changes programmatic will bring to so-called
traditional media channels. Media sales as a discipline is changing, and
those changes are important and durable. With every step taken down the
programmatic path, a seller is putting in to place the principles, processes,
and practices that will guide their business for the next several decades. As
such, their programmatic solution provider should have a similarly advanced
long-term vision.

29
With a draft operating model and an understanding of the
buyer and seller tools; the next step is to consider the
various types of media inventory available and how they
can be accessed programmatically.
There are various inventory ‘types’, IAB Europe and IAB
US define four key types of inventory as outlined below.
1. Automated Guaranteed
Traditionally, programmatic buying is used to source demand when a media
owner has not sold it directly already. . However, ‘automated guaranteed’ or
‘reserved’ complements direct sales, and is trafficked alongside direct demand
in the publisher’s ad server via the automated reserved technology solution.
Therefore, the media cost of automated reserved campaigns is aligned with
direct sales CPMs, and is frequently higher due to the increased value
generated from the programmatic transaction. The buyer and seller have a
direct relationship, therefore both parties have greater transparency than is
possible in open market real-tie bidding (RTB). Conceptually, an automated
reserved buyer should be able to negotiate for all inventory a media owner
has available for fulfillment in its publisher ad server. In reality, today’s
inventory availability is lacking due to tight integrations between the
automated reserved technologies and other systems in the programmatic
stack.
2. Unreserved Fixed Rate
An unreserved fixed rate transaction is typically facilitated by a media owner’s
supply side platform (SSP). An SSP generally manages all non-direct sold
demand from a publisher, typically at a lower priority, and rate, than these
campaigns in the publisher’s primary ad server. Therefore, media costs for
unreserved fix rate transactions often fall beneath those of automated
reserved buys.
The media owner and buyer transparency is still very high compared to other
models, but the addition of the SSP creates dependency on the capability of
this platform to surface information about the transaction to both parties. Due
to the SSP’s position in the publisher ad server prioritisation queue, potential
inventory availability is limited by the media owner’s openness to send traffic
to the platform. It should be noted that automated reserved solutions are in
their early stages, compared to unreserved fixed rate deals, meaning that the
2.4 Inventory
and Sourcing
Methods

30
inventory volume transacting over unreserved fixed rate currently exceeds that
of automated reserved.
3. Invitation-Only Auction
Typically, an invitation-only transaction is facilitated by a media owner’s Sell
Side platform (SSP). Because an SSP generally manages all non-direct sold
demand from a publisher it is typically at a lower priority, and rate, than these
campaigns in the publisher’s primary ad server. Additionally, because
unreserved fixed-rate transactions are usually associated with a volume
commitment from a buyer, a media owner usually prioritises unreserved fixed
rate demand above invitation-only auctions. However, this is not always the
case, as frequently media owners will market a ‘first look’ invitation-only
auction.
Usually this tactic only works on very high quality, owned and operated
inventory. The level of transparency is very high, but the auction dynamics in
SSPs create complexity in terms of surfacing inventory and pricing information
to both buyers and sellers that may or may not be realised by each platform.
Potential inventory availability is limited by both the position of the invitation-
only auction within the SSP priority system and the placement of the SSP in
the publisher’s ad server. However, because an invitation-only auction
includes many buyers, where an unreserved fixed rate transaction includes
just one, more inventory is exposed to this transaction model than unreserved
fixed rate at present.
4. Open Auction
An open auction transaction is usually facilitated by an SSP, with a lower
priority than demand sold directly in a publisher’s ad server. The open auction
includes almost all buyers, so the prioritisation will be lower than the other
models, which by definition have more direct relationships between the buyers
and sellers. Since the open auction is usually the lowest priority, the lowest
rates are charged as, generally, some demand is better than no demand and
the level of transparency is very high.
However, the inherently indirect relationship between buyers and sellers in the
open market means transparency into inventory and the number of
intermediaries in a transaction is dependent on the reporting capabilities for
demand side platforms (DSP) and SSPs involved. Potential inventory
availability is the lowest of all models, due to the priority position within both
publisher ad server and SSP systems. The open market receives ‘what is left’
after direct sales, automated reserved, unreserved fixed rated and invite-only
auctions. However, since adoption of programmatic transaction models varies

31
widely across media owner and market, the open market has the most
programmatically enabled supply today.
The Future: Potential and Reality Converge
Despite industry predictions that the majority of inventory will be traded
programmatically within the next few years, there remains in many markets a
gap between the potential and the reality. Technology and commercial models
support full programmatic availability of all digital media and the efficiency and
effectiveness of these tools is improving rapidly. However, media owners with
strong businesses built on direct sales are wisely managing the evolution to
programmatic media sourcing with some caution. The shift requires strong
technology capability, third party integrations and evolved internal sales and
ad operations human capital.
As media owners differ in their approaches to build, buy or partner these
functional areas, we see varying deltas between what’s ‘available’ and what
‘could be available’ in programmatic buying today. However, buyers such as
Interpublic Group have declared goals of 50% of all digital media being
procured programmatically by 20174
, that delta looks like narrowing quickly as
adoption and sophistication of the programmatic technology evolves.
Market Differences
The following outlines some key differences between developing and mature
markets.
Developing Markets: Due to relative market maturity and lower levels of
buyer/seller confidence, these tend to be more ‘closed’ markets, with less
transparency provided by sellers in their real-time bidding (RTB) activity. For
more ‘premium’ buyers, transparency is enabled via deal IDs rather than being
broadcast within RTB ad calls.
 Regarding sales practices, there is greater emphasis on cost per click
(CPC) and performance budgets, representing both the role of
programmatic to the seller (remnant) and the mix of buyers in the
market (fewer brand / premium players).
4 IPG to Automate 50% Of Media Buying, Expand Programmatic, MediaDailyNews:
http://www.mediapost.com/publications/article/205836/ipg-to-automate-50-of-media-buying-
expand-progra.html

32
 Within the publisher yield stack, programmatic is often used as a
remnant filler, with little upstream integration into the main direct sales
pipeline.
 General access to inventory is more limited, and much of what is put
out into the market is ‘blind’ or ‘semi-blind’ – meaning little information
about the publisher’s site and specific placement attributes is shared.
However, conditions and practices are changing rapidly. Developing markets
are looking to more mature regions, learning and adapting quickly to premium
programmatic sales practices.
Mature Markets: The activities of more mature programmatic markets very
much reflect the beginnings of what programmatic can mean for digital media
overall in the future. Increasingly, publishers are relying on a single system to
manage and optimise all sales – balancing the various sales channels through
which buyers are sourcing inventory – direct sales, programmatic direct,
private marketplaces (PMPs), deal IDs, ad networks, demand side platforms
(DSPs), trading desks, etc.
Given the direct integration that occurs between the supply side platform
(SSP) and demand side platform (DSP), these constitute ‘invitation-based’
relationships. With that level of integration, depending on SSP capabilities,
programmatic premium sales can be managed. With that comes: a
considerable level of product transparency; specific inventory targets and
guarantees; more precise negotiated prices and/or publisher price controls.
Again, depending on the capabilities of the SSP, these kinds of direct
integrations give buyer and seller greater flexibility, as the line between
negotiated IO-based deals and programmatic begins to blur. The seller is able
to optimise holistically, regardless of sales channel, and the buyer is able to
access the inventory on offer however they see fit (whether through RTB at
one end of the spectrum, or a faxed IO at the other end of the spectrum). In
some cases, in fact, some publishers are allowing correct booking access to
some or all of their advertising inventory through the intermediary capabilities
of an SSP.
For both buyer and seller, this more mature programmatic market state carries
two primary benefits, not to be underestimated – operational efficiency and
higher returns:
 For the media seller, this gives them access to the widest possible set
of buyers and the most diverse mix of demand. With the right

33
supply/demand optimisation, that translates directly to the highest
possible sales revenue and inventory yield.
 For the media buyer, they in turn have a constant view into the
broadest possible supply and (virtually) infinite options to overlay user
data to reach the consumers they are after. This translates to more
accuracy in hitting their target audience; more control over reach and
frequency objectives; and ultimately, higher campaign return on
investment (ROI).
As noted in section 2.2 single media schedules are a
thing of the past and today advertisers need to use an
array of media channels to connect with their targets
groups and create consumer touchpoints. Audience data
therefore needs to be more consistent across all channels
to make reliable comparisons between channels as well
as to evaluate the net results achieved across all
channels used.
IAB Europe has formulated a set of goals and
recommended actions for online audience and ad
effectiveness metrics and KPIs to attract further brand
advertising investment into digital channels across
Europe. This IAB Europe Measurement Blueprint
5
includes recommendations for cross-channel
comparisons.
Key recommendations outlined in the Blueprint on deliverable metrics are:
 Online audience measurement should integrate with and enhance
other established metrics, such as net reach, frequency and GRP
levels
 Online audience metrics should be reported individually by
PC/Laptops, Tablets, Smartphone devices and Connected TVs both
for web or in-app
 Audience metrics should be combined whenever feasible with data
identifying the impact on achieved brand KPIs, such as awareness,
brand affinity, and purchase intent with the aim, for example, to
5 IAB Europe Blueprint on Online Audience Measurement and Ad Effectiveness Metrics and
KPIs - http://www.iabeurope.eu/research-and-papers/iab-europe-blueprint-online-audience-
measurement-and-ad-effe
2.5 Thinking
Cross-Channel
and Cross-
Device in a
Programmatic
World

34
measure and describe the fuller consumer journey across the media
landscape
 Results should primarily be compatible with other domestic media
currencies, and secondly with online audience surveys across markets
o For example, target definitions should be compatible by
demographic and/or lifestyle segments and have consistent survey
universe definitions (and weighting) across media channels
 Exposure measurement should take into account the viewability of
display or video ads to deliver realistic opportunity to see (OTS)
outcomes
The Role of Identification
As digital advertising increases its ability to deliver a more relevant experience
based on interests and preferences, the hidden foundations of audience
identification become increasingly apparent. Identification (and IDs) underpin
the capabilities of this more evolved digital ecosystem, by providing the
lynchpins to initiate, link and customise interactions along the consumer
journey. Unfortunately, we’re no longer in a world where the desktop web
cookie will suffice.
Now the consumer journey is an interwoven track of touchpoints that occur
across numerous channels (including web, app, social, email, in-store and
offline) and devices (desktop, mobile, and now the internet of things). For
example, without a common means of identifying audiences on mobile web
and mobile app, it is not possible to understand how the audience interacts
with at different touchpoints which limits the potential for providing a seamless
consumer experience.
The challenge for those thinking cross-channel and cross-device is to work
around the industry’s infamous ‘data silos. That means understanding
identification in the context of EU privacy law, and the technological solutions
on offer to a sufficient level to be able to appropriately consider in-house
identification strategies and the right complementary vendor solutions.
Cross-channel and cross-device programmatic strategies rely on two
complementary approaches to identification. Deterministic device IDs are
based on a one-to-one relationship with the user (like cookies and device IDs),
offering accuracy and longevity, but lacking in coverage and commonality.
Statistical device IDs work off the statistical probability of uniqueness for any
single device, and are typically used for linking audience activity across
channels and marketing technology players (where similarities in a user’s

35
device characteristics can be recognised) and across household devices
(based primarily on IP address).
However far advertisers and publishers wish to take their own identification
resolution strategy, they should have visibility of those of their partners.
Moving on from Last-Click Attribution to Reflect the Multi-Screen
Consumer Journey
In the interwoven journey, each touchpoint makes its own contribution to make
to the final conversion, but the simple metrics of last-click attribution – on
which much of the industry still relies – are insufficient to measure this impact.
Last-click attribution only recognises the final touchpoint before conversion,
and this approach typically allocates less importance to high-funnel activity
such as prospecting than to lower-funnel activity such as retargeting.
Engaging consumers throughout the entire journey – before, during, and after
purchase – should be the goal of any marketing campaign. By recognising the
unique contribution of every touchpoint regardless of channel, advanced
attribution makes this a real possibility.
Embracing a more advanced form of attribution that determines the
proportional credit each individual touchpoint deserves has a number of
benefits for marketers. Firstly it can drive brand awareness and consumer
engagement by helping marketers to understand the influence of each
marketing activity on the consumer’s decision to purchase. Secondly
advanced attribution can provide an overview of the whole marketing
ecosystem, allowing marketers to understand the interaction and impact of
touchpoints between distinct channels – for example, the effect online
advertising has on offline sales. Finally advanced attribution can be used to
inform real-time decision-making and media buying by understanding the
value of each impression.
Different types of advertising are more suited than others to engaging
consumers and driving action at different stages of the sales funnel. Video
advertising, for example, is a powerful tool for increasing brand awareness
among consumers who have never encountered a brand before. It is a high-
impact format that familiarises consumers with a brand creating early-stage
demand. Alternatively, mobile is the ideal channel for reaching consumers at
the mid-funnel product research stage, as smartphones are often used to
compare and review products. Finally, social is useful for maintaining a close
relationship with existing consumers and building brand loyalty by offering
special offers or early access to new products.

36
Advanced attribution creates a synergy between interactions in different
channels and means marketers can deliver on the promise of a seamless
consumer experience from channel to channel and device to device, both
online and offline. Today’s marketers can benefit from taking a single-platform
marketing operating system approach to the consumer journey. Marketing
technology empowers marketers to adopt a truly cross-channel strategy,
remove silos between channels, and optimise every interaction across all
addressable media to deliver finely targeted ad messages to consumers at all
stages of the path to purchase.
Programmatic Makes a Cross-Channel Approach even More Important
by Rewarding Ads that Really Work
Programmatic is a fundamental evolution in the approach marketers take to
optimise budgets across the whole advertising ecosystem. Programmatic
automates real-time decision-making and media buying, assessing the value
of each particular impression based on its role in the conversion path and
determining whether to bid and by how much. Put simply, programmatic
rewards ads that work and allocates budget to high-performing impressions,
meaning it is vital that marketers can measure the performance of ads across
all channels, devices, and levels of the sales funnel.
Combining attribution technology with programmatic can provide marketers
with a solution to cross-channel measurement. This approach allows all digital
touchpoints to be mapped out and assigned a value based on the impact that
each instance of brand exposure has on consumer behaviour. It can have a
significant impact on cross-channel success by informing each and every bid
decision, increasing efficiency by bidding on impressions that are likely to lead
to the desired outcome – whether performance or brand related. This ensures
that the correct marketing mix is being directed to the right consumer.
Marketers that wish to benefit from better insight into the ROI of campaigns
across channels, improve results, and justify greater marketing budgets will
benefit from adopting a single-platform marketing operating system that
simplifies the often complex processes faced by marketers.
Future Trends
Over the coming year the focus will shift from the inefficiency of last-click
attribution towards the development of more sophisticated attribution models
that can measure the impact of impressions across all channels. Selecting the
right technology is key in switching from last-click to advanced attribution. The

37
key for marketers will be developing multi-platform attribution models that
align with their unique business goals, and programmatic platforms will need
to be flexible enough to work with any advanced attribution partner.
While several of the major technology providers are creating their own
advanced attribution solutions, these may not be the best option for all
marketers. These types of companies leverage the power of their own data
and data stack, and as a result can be seen as inflexible in their approach,
with a focus on their own ‘walled garden’ environment and little compatibility
with the open web.
Independent technology providers, on the other hand, are able to work
agnostically across the entire ecosystem, providing solutions that can be
customised to marketers’ individual attribution needs. Marketers will
increasingly utilise marketing operating systems to leverage advanced
attribution data to efficiently inform every bid decision and spend. From one
central platform, marketing operating systems allow marketers to implement
custom attribution models into real-time decision making, understand the
proportional impact of different media pathways, and quantify the relative
contribution of different marketing tactics.
Advances in cross-device tracking and targeting will also play a key role in the
development of attribution. We will begin to see the cookie gradually being
replaced by new targeting solutions from large media companies, publishers,
exchanges, and third-party providers. Bridging technology will emerge to
seamlessly integrate distinct solutions for the purposes of targeting,
attribution, and analytics – recognising users across cookie or cookie-less
channels and devices.

38
In section 2 this White Paper touched on the value of a Data Management
Platform (DMP) for the buyer, but data is increasingly important to all parties in
the digital advertising industry. As marketers start to use data to drive efficiency
and effectiveness, they increasingly value publishers’ data. This data gives
context, audience insight and thus drives the value of audience impression. But
data, whether big or otherwise, needs to be gathered and managed carefully to
deliver true value. In this section we look at where data comes from, how it is
being used and what issues this raises for those considering programmatic
models.
It is important to note that where personal data is collected and processed, the
data controller will need to obtain consent, a contract or clearly identified
legitimate interest in order to legally process the data. All parties should ensure
that they adhere to EU privacy law.
First we should define some data terms:
First Party Data
Advertisers have transactional CRM data, whether online or offline, from
people who have made purchases at physical stores or via an e-commerce
site; from sales leads; and from people who have interacted with their call
centre. Advertisers also have data regarding user behaviour on their website
and campaign performance data from email, display, video, mobile, direct
mail, TV, print, and other media. First party data is often gathered through tag
management where small data files are put on each web page or email as a
pixel. This pixel tracks user interaction and when synchronised in real-time
with the DMP provides a new level of visibility of activity on the site. Once this
data has been gathered, the system allows a marketer to transform the
information by segmenting transaction values from a shopping cart into a set
of marketer-defined boxes. Transforming the incoming data enables the later
processes of segmentation and targeting to be faster and easier, by
standardising the raw event data into more usable information to begin with.
Taken together, first party data represents a massive amount of marketing
information that’s frequently left untapped but in programmatic, it is often the
starting point for gaining insight into ideal audiences.
3 Data

39
Publishers have their own first party data and can use this to create a
personalised experience for each and every user – ensuring users receive a
personally relevant and meaningful experience. To do this, it’s important to
understand what a user is coming to a publisher site for: what are the tasks
they are trying to achieve, which products they are engaging with, and what
that means in terms of their interest. From a publisher (or media owner)
perspective, they look at all
Publisher’s take into consideration many facets of user’s behaviour; their
search activity on the site, site content, social interaction e.g. blogs and likes,
publisher apps, engagement with ads on the site and publisher emails.
Publisher’s asses the relevance of the different interactions in order to deliver
a personalised experience for the user, for example surfacing content that is
more likely to engage.
Second Party Data
Second party data is making use of someone else’s first party data, for
example a publisher’s audience data. Second party data enables a marketer to
augment their first party data with a new data set, enabling the advertisers to
enrich the knowledge about their audience at scale.
Second party data has strong advantages, although it may not drive
performance at the same rate as retargeting or first party data. Second party
data also provides information about audience behaviour and the consumer
journey, for example if a user is in the top decile of users visiting a sports
website over the last two weeks, the advertiser would know they are really a
sports fan – they would not just be someone who stumbled across a sports
article. Similarly, someone who visits a search engine and searches for ‘buy
new HDTV’ is clearly in market for a new TV with the intent to purchase.
So publisher data can be classified in a number ways:
 Interest level data: a user is interested in sports, music,
entertainments)
 Intent level data: a user has demonstrated their purchase intent, they
are in market now for a specific product)
 Other audience data: registered demographics, geographics
 Demographic data: this predominantly comes from users as they
register with the publisher
 Search retargeting data: users who have searched for a particular
keyword or set of keywords
 Mail insights targeting: users who have received emails from specific
commercial domains.

40
 App usage data: users who over-index on engagement with certain
apps; not only if a user has downloaded that app, but how and when
are they actually using it
 Social data: users who have liked, shared or reblogged specific
brands or subject areas.
This data might be collected in a number of ways:
 Declared (through a registration process)
 Direct (through actions the users has carried out)
 Implied / Inferred (from actions that a user has carried out)
Third Party Data
In-house information is important, but its value increases dramatically when
blended with audience data from third party providers. With a data
management platform, first party data helps to highlight valuable audiences,
then third party data can be overlaid to develop an understanding of attributes,
behaviours, and content consumption. For example, a DMP can show, from
second party data, where visitors engaged or converted, and from third party
data helps to illustrate factors such as economics, family position and stage of
life.
How Advertisers can use Data Effectively
Today CMOs are using data visualisation tools to better understand the
consumer journey. They’re more directly in tune with campaign performance.
Today’s tools and techniques mean that advertisers don’t need to be a
technical expert in order to create a more rewarding consumer experience.
Data management platforms hold a wealth of anonymous data about prospects
and consumers. This information can be used to deliver a more tailored
message, one that engages the audience precisely because it is relevant and
useful. Relevance starts with smart audience segments, so advertisers can
target consumers with messages and promotions that are more likely to
engage, and therefore more likely to convert. Advertisers can also look at where
audiences are in their consumer life cycle to determine what to say to them,
through which channel, and when. At the same time, advertisers can make
data-driven decisions about exactly where to place those ads, optimising the
allocation of their advertising spend.
For example, as a mobile network carrier, an advertiser might want to reach its
opted-in subscribers with renewal messages across multiple channels. A DMP
can help to segment these ‘renewal’ consumers further, perhaps based on how

41
long they’ve been with the carrier. Or a DMP can enable you to access second
and third party data to help to promote smartphones to relevant audiences
such as gadget lovers. Thinking beyond display ads and online video, a DMP
can be used to optimise websites for individual visitors. A home page should
be dynamic, with content that’s tailored to what a centralised database knows
about returning and new consumers. And as visitors browse a site or sees one
of an advertiser’s ads on another site, the brand has the data required to define
a consistent, targeted experience across all communications.
Campaign Insights
When using data to build better consumer relationships, the opportunities for
insight are significant. As marketing campaigns are executed, across both
online and offline channels, the results can be added to the data ‘story’. Each
campaign adds substance to the story and instructions for the next effort,
creating an upward spiral of better and better performance. Of course, even
underperforming messages or audiences can provide guidance for future
campaigns.
As data-driven marketing efforts mature, marketers need to make sure to
analyse this data story and its associated costs. Although cost per impression
continues to decline, thanks to an exploding supply of ad inventory, the cost of
audience data is not following suit. When an advertiser executes a look-alike
campaign, where they find new audiences that fit the profile of good
consumers, the data management platform can report the data cost as a
percentage of cost per impression, enabling the advertiser to see which third
party data providers are driving the most value. They can then optimise their
next campaign by culling out the data providers that did not perform well. The
DMP should show the return on investment (ROI) of the whole campaign, as
well as the ROI of each of the smaller segments in the campaign. Some
surprising results may occur, like a data provider that yielded a lot of
impressions but a low ROI. This information can be used to create better
performing audiences for the next campaign.
Understanding the Consumer Journey
In today’s multi-screen world, the odds that two consumers will follow the
same route to purchase from an initial brand encounter are slim. However,
even though communication pathways have grown more sophisticated, the
consumer journey hasn’t really changed all that much. Consumers still pass
through four key stages: brand awareness, consideration, intent, and,
assuming all goes well, conversion.

42
Today, thanks to the value and efficiency provided by programmatic
marketing, marketers are able to design strategies for countless possible
outcomes, and to deliver the best possible results whatever path their
consumers follow. By understanding how each stage along the consumer
journey functions, and by using the appropriate programmatic tactics and
benchmarks to reach desired audiences at each stage, the smart marketer is
better equipped than ever to deliver optimal results.
 Awareness: At this stage, the advertiser’s primary goals might include
introducing audiences to the brand, increasing site traffic, or finding
prospective consumers with some purchase intent. For most brand
awareness campaigns, the desired audience will be quite large. With
the availability of high-quality inventory, new ways to protect the brand,
and the technology to optimise campaigns toward specific goals, real-
time bidding (RTB) has become an efficient way to engage consumers
through this initial awareness stage, providing access to the reach and
scale needed to drive awareness. Tactics such as look-alike modelling,
which allows an advertiser to expand its target audience based on
characteristics of converted consumers, can be useful. Supporting
such efforts with a brand-lift survey can help to highlight which
audiences found the brand most favourable, pointing to the most
valuable impressions and improving the quality of audience
segmentation.
 Consideration: Narrowing the audience segments with broader
branding tactics used during the awareness stage can help to increase
consumer consideration. Content, predictive, and behavioural targeting
can be used to enhance audience segments. Content targeting
matches the content of an ad to the content of a web page the
audience is viewing for more relevant messaging. Predictive targeting
during the consideration stage will allow marketers to refine their
desired audience based on data gathered during the awareness stage;
behavioural targeting will help to highlight which audiences are most
likely to show interest in the brand.
 Intent: By signalling interest in a brand, with site visits or other
advanced actions, prospects can be categorised in the intent stage.
The goal during this stage is to turn interest in a brand into desire for
purchase. To achieve this, a combination of behavioural targeting and
retargeting can be used. Retargeting works by delivering ads in
succession, which serve as reminders that greatly increase the
chances of a conversion. While predictive targeting, behavioural
targeting, and retargeting are all effective tactics individually, a strategy

43
that combines all three will yield the best results, helping a marketer to
nurture the consumer relationship.
 Conversion: Finally, the audience reaches the last and most important
stage. Once a conversion is made, real-time bidding allows audiences
to be appropriately segmented and deploy relevant follow-on
advertising. Using multi-touch attribution (MTA), which brings together
information from the consumer journey, marketers are better able to
understand how and why consumers engaged with their brand along
the journey, and thereby refine their audiences and strategies and
implement tactics such as behavioural targeting, content targeting,
look-alike modelling, site extension and dynamic creative optimisation
(DCO).
Understanding Data to Drive New Consumers
Marketers can use look-alike modelling, a data management platform (DMP) to
employ third party data to discover audiences with similar attributes as existing
consumers, enabling an advertiser to expand its reach effectively.
As a DMP finds more compatible segments, the marketer ends up with an
audience sizable enough to execute against in digital marketing efforts. Look-
alike modelling works best if the DMP has a large ‘digital footprint’. A platform
that has pre-populated more third party data providers will uncover more look-
alikes, giving the marketer better, more profitable options when it comes to
purchasing media.
Five Data Points every Advertiser should look at
Over the past ten years, digital advertising has matured from an imprecise
practice of ‘seeing what works’ to a highly targeted system of matching the
right audience with the right ad. Below are five tips for using data to build more
profitable consumer relationships:
1. A data management platform (DMP) offers greater potential than siloed
systems like CRM to understanding the consumer information that
already exists across the organisation. As it is consolidated into one
place it is important to make sense of it and gain new confidence in
marketing decisions. It remains to be seen whether this is a standalone
DMP or part of an integrated ‘marketing cloud’.
2. Third party data can be blended with first party in-house and second
party publisher data to provide a complete picture of an advertiser’s

44
consumers, answering a wide range of questions about audience
attributes and behaviours.
3. Smart audience segmentation can help to improve marketing
performance. Consumer relationships can be enhanced by delivering
tailored, consistent messages across all channels.
4. Look-alike modelling is a method of finding audiences who share
attributes with an advertiser’s existing consumers or other desired
audiences. Understanding of audience segmentation can pave the way
for expanding marketing reach.
5. Data-driven actions deliver fact-based insights. At the same time,
marketers want to keep an eye on data costs to make sure they’re
seeing the results required.
A more accurate understanding of audience can drive improvements across all
campaigns—including TV and print. While marketing intuition will never be
obsolete, marketers can also rely on a modern marketing platform to guide
strategies and actions.
Look at Data Beyond the Click
While real-time bidding (RTB) mechanics are reminiscent of those of the
search auction, it’s important to remember that display campaigns may not get
the same number of clicks as search campaigns. In a given month, only 16
percent of people click on display ads6
even though they may be exposed to
multiple display ads in a single online session. The difference with display is
that this low percentage of clickers doesn't have anything to do with display
ads' effectiveness, it simply indicates that clicks are not the best measure of
value for these ads. When people visit content sites, they’re actively engaged
and don’t intend to click away to visit another site. This means that even an
interesting display ad may not be enough for someone to click away
immediately. Instead of clicking in the moment, a person who sees an ad will
often search for the product the ad mentioned or visit an advertiser’s website
later on. Display ads simply take longer to have an impact as consumers
respond to them at their own pace.
6 Gian Fulgoni, “Who Will Rid Us of This Meddlesome Click.” -
http://www.comscore.com/Insights/Blog/Who_Will_Rid_Us_of_this_Meddlesome_Click

45
Using Programmatic Technology to Evaluate the Metrics that Matter
Understanding Audience
This chart allows an advertiser to understand who their audience is and
tailor offers and promotions to the relevant audience, and understand
which audience responded to their media campaigns.

46
Understanding Performance
Real-time data can now be used to understand key performance
indicators. The chart on the left shows delivery of the campaign against
a specific budget as well as performance by day against the campaign
goals, where that is cost per action (CPA), return on ad spend (ROAS)
or click through rate (CTR).

47
Prospecting Insights
This chart allows the advertiser to clearly define which consumers are
genuinely new to their site (new prospects) and which have already
visited the site in the past. Marketers can track how many new
consumers were targeted, how many of them went on to visit the site
and then eventually converters. This can then be compared to the
average site conversion to show lift.

48
Time of Day
Through real-time tracking and measurement advertisers are able to
see exactly what day and what time of day the campaigns drive the
most consumers to convert. This also provides the advertiser with
valuable information about their consumers and when they are most
receptive to advertising, which can help inform their media planning.

49
Time to Convert
 When a user is first exposed to a display ad,
 How many ads it takes to drive them through consideration to
visit the advertisers site,
 What percentage of those users convert as soon as they visit
the site
 What percentage leave the website and need to be exposed to
several more ads in order to return to the website and
complete the action. This can help advertisers define their
advertising cookie windows, frequency and split of prospecting
activity versus retargeting activity in order to deliver the most
effective ad campaigns.

50
Questions to Consider when Evaluating a Successful Campaign
Understanding the consumer journey is critical to optimising display campaign
performance. While looking at conversions and site visits will help to unlock
more pieces of the conversion puzzle, there are also other factors that should
be considered.
Below is a list of questions that marketer’s should consider in order to
understand whether they are getting the most out of their display advertising
campaigns:
Questions to ask Why it matters Metrics and insights to
consider
Am I reaching new
prospects with my
campaign?
Expanding audience is critical
for driving incremental
conversions to ultimately
grow business.
Prospecting Reach Rate
Measures the percentage of
new people reached
Am I reaching the
right audience?
It’s easy to reach a lot of
people with cheap ad
inventory or other spam
tactics, but this most likely a
waste of marketing budget.
Prospecting Visit Rate
To see if a good job of
targeting the right people
has been done, measure
how many new prospects
visited the site.
Are my prospects
qualified?
If the prospects are qualified,
they are more likely to
convert on the site.
Measuring the rate of
visitors to converters will
inform if the marketer did a
good job of qualifying their
prospects.
Do my prospect
require retargeting?
If site visitors don’t convert on
their first visit, there is an
opportunity to reach them
again with retargeting. Do
you know your visitors’
behaviour and how much
retargeting you should be
doing?
First Visit Conversion
Ratio
Of all the people who
converted, what percentage
converted in one visit
versus multiple visits? A low
percentage indicates that
you should increase your
retargeting efforts, and a
high percentage indicates
that you should reduce your
retargeting efforts.

51
How long did it take
my prospects to
convert?
The period of time it takes
from a prospect’s first
impression all the way to
conversion is called a
‘lookback window’. If it took
most people 20 days to
convert then the marketer
should set its lookback
window to at least 20 days, or
even longer.
Time to Convert
The period of time between
a consumer’s first
touchpoint with the ad
campaign(s) all the way up
to the actual conversion.
How many
touchpoints were
needed?
To maximise efficiency and
help determine ad frequency,
look at how many ads were
served before and after the
site visit. There’s no need to
run 10 ads when you only
two are needed.
Total Touchpoints from
First Touch
Determine the number of
impressions that were
served per user before the
conversion event, before
the site visit (prospecting)
and after the site visit
(retargeting).

52
The path to programmatic brings greater efficacy and
efficiency to online advertising, but for true impact there
must also be due consideration of the quality of media
being transacted upon. In programmatic trading, media
quality is a concern to both the buy and sell side. For the
sell side, there is a need to offer high quality inventory to
advertisers increasingly looking for quality marketplaces.
While for the buy side, the concern is that their
advertising is placed in brand safe, viewable
environments that are free from fraudulent (non-human)
traffic.
The diagram below outlines the four quality segments that IAB Europe has
identified within its Brand Advertising Committee.
Diagram 1 – IAB Europe’s Quality Topics
The following section explores the topics within the delivery segment including
viewable ad impressions, non-human traffic and brand safety. IAB Europe is
working with the World Federation of Advertisers (WFA) and the European
4.1
Quality
4 Further Strategic Considerations

53
Association of Communications Agencies (EACA), as previously mentioned in
section 2.3 to address these topics on a European level.
Viewability
Viewability is a key component of media quality and with the need to shift
advertising currency from served impressions to viewable impressions;
viewability is quickly becoming a key factor in determining quality.
IAB Europe’s Measurement Blueprint notes that ‘definitions used for key
metrics, such as exposure or impressions need to be tightened to avoid
eroding advertisers’ confidence in the data. Measuring the viewability of
served impressions will be a vital step towards delivering realistic exposure
levels. Specific efforts should be made to overcome any technical difficulties
measuring the viewability of display or video ads.’
As greater programmatic spend moves into mobile and video formats it is
becoming increasingly important to be able to verify and measure more
aspects of a media plan. There is a need to be able to offer viewability
reporting across display, mobile and video to allow like-for-like comparison
across multiple channels.
The Media Rating Council (MRC) in the US first lifted its advisory on
transacting on viewable impressions in 2014 and the organisation has
released standards for viewability measurement. Whilst standards and
guidelines are now in place for display, mobile, rich media and video, there
remains discussion as to what defines viewability in the marketplace and how
it is measured. IAB Europe has outlined the European position in the Viewable
Impressions White Paper7
published in February 2015.
Viewability is defined as an ad having the opportunity to be seen by a human.
Measuring viewability depends on more than simply the location of the ad and
how much of it is in view. It is also dependent on the timing of the viewing. To
tackle location, time and size, two methodologies are available:
 Page Geometry - assesses the position of the ad relative to the
viewer’s screen
 Browser Optimisation - analyses changes in ad processing speed
when the ad is on or off the screen.
7
IAB Europe Viewable Impressions White Paper (February, 2015) -
http://www.iabeurope.eu/research-and-papers/iab-europe-viewable-impressions-white-paper

54
There is growing demand for up to 100% of traded programmatic media to be
viewable but significant measurement challenges remain, particularly in video
and mobile. The limitations of current technology prevent all served
impressions from being fully measurable. For example, many high impact
placements, such as page takeovers and other custom placements are still
not consistently measurable. So it must also be taken into consideration that
unmeasured impressions should not be assumed to be non-viewable
impressions.
Another factor to consider when looking to improve media quality through
viewability, is that measurement vendors produce campaign level viewability
data that can vary between vendors due to different measurement
methodologies. If buyers and sellers in the programmatic space agree on a
single measurement vendor ahead of time, they are likely to avoid costly,
labour-intensive, manual processes of reconciling different sets of viewability
data. Agreeing on a single vendor ensures greater accuracy in reporting and
eliminates the need to reconcile data from multiple sources.
Non-Human Traffic
Viewability is a key step towards attaining quality media. Another step is to
exclude fraudulent inventory (non-human traffic). Depending on the inventory
source, nearly half of all impressions are not in view.
Non-human traffic is a costly issue affecting the entire online advertising
industry, but recent industry proactivity, like the efforts of the JICWEBS led
anti-fraud working group in the UK, and the wide availability of tools to
manage the threat have greatly mitigated the risk.
Within the programmatic space, ad fraud is widely discussed and the
perception of the presence of non-human traffic far exceeds the reality.
Firstly, what counts as fraudulent activity needs to be clarified. Fraudulent
activity can be defined as the deliberate practice of attempting to serve ads
that have no potential to be viewed by a human user, which typically online
this is carried out through the use of bots. The IAB in the US has laid out its
view on what comprises fraudulent activity in their Anti-fraud Principles and
Taxonomy8
.
8 IAB Anti-Fraud Principles and Proposed Taxonomy -
http://www.iab.net/media/file/IAB_Anti_Fraud_Principles_and_Taxonomy.pdf

55
There are effective methods to identify fraudulent activity in online campaigns
and inventory. By applying both big data and session-based signal analyses,
anti-fraud technologies can use historical learning and real-time signals to
make a decision about the existence of non-human traffic on a given web
page.
For traders dealing via programmatic channels, there are anti-fraud targeting
segments that can be applied to campaigns that target away from fraudulent
inventory before the ad buy/sell transaction takes place. Agencies and
marketers can further exclude fraud from their programmatic campaigns by
leveraging integrated blocking technology that stop ads from ever serving on
fraudulent pages.
Although the industry is working hard to eliminate fraud from all sides of online
advertising, detecting whether traffic is fraudulent is not straightforward, and
fraudsters are continually innovating and finding new ways to prevent
detection.
Brand Safety
After applying pre-bid targeting segments for viewability, and ensuring to block
for ad fraud, programmatic spend will already be more effective and the
quality of media far improved. However, even a fraud-free campaign with
excellent viewability can be compromised by problems, such as unsavoury
images and inappropriate content. To run the most effective campaigns whilst
ensuring best media quality, there is a need to think beyond viewability and
non-human traffic, and to address brand safety factors as well.
Every advertiser will have its own interpretation of the type of content they
deem appropriate or not.
Due to its subjectivity, brand safety and the tools available to address it need
to be customisable to fit a brand’s requirements. This subjectivity in brand
safety means that as an industry we are unable to reach a consensus on a
brand safety standard against which the market can operate.
Campaign targeting can help to reduce the risk of advertising appearing
alongside inappropriately aligned content that can result in off-brand
associations and even regulatory concerns. Brand safety blocking and pre-bid
targeting can be customised by content category (violence, illegal downloads,

56
alcohol content, hate speech, offensive language, illegal drugs or adult
content) as well as risk threshold (low, moderate, high risk).
Summary
The potential of programmatic advertising can only be maximised when a
comprehensive approach is taken to address media quality - viewability, non-
human traffic and brand safety are intrinsically linked, but they are still only
parts of a larger media quality equation, as outlined in diagram 1 on page 47.
Attention must be paid to the sequence in which these issues are managed
and the tools that are used. The current situation is such that variations in
these factors can substantially affect the measured outcome and perceptions
of campaign performance. Those trying to manage media quality are well
advised to work closely with their technology and media partners, and are
encouraged to make it a search for agreed definitions and measurement
criteria, rather than an absolute truth. After all, the true power of content
verification and viewability comes not from being ‘right’ but from being able to
work with partners to improve campaign performance and this often means
collaborating around the same standards.
Mobile advertising has several unique capabilities that lend
themselves well to reaching and engaging with relevant
audiences to enable advertisers to achieve their objectives.
Increasingly, challenges around the difficulties in executing
campaigns at scale with appropriate tracking are viewed as
being surmountable, and the scale that can be achieved
makes mobile an unmissable opportunity for any advertiser
wanting to engage with its target audiences.
Here are some considerations for marketers to reach and engage audiences
in the increasingly mobile world:
1. Audience Profiling
The beauty of mobile devices for marketers is that they can gain insight into
how people move throughout the day and can measure consumer-brand
engagements in more detail than ever before. For example, noticing someone
appearing at airports on Mondays and Fridays on a regular basis increases
the likelihood that they are a business traveller.
4.2
Programmatic
Mobile

57
Mobile devices come loaded with capabilities that enable much more involved
experiences. The list includes a touch screen, GPS, an accelerometer, a
gyroscope, a compass, a camera and a microphone. Between them, these
allow an advert to understand where someone is, what direction they are
facing, what angle the phone is being held at and whether they are moving or
not. Plus audio and visual input can be captured, and the experience can be
controlled using touch as opposed to a keyboard or a mouse. There are many
great examples of ads that have provided real utility to users, and therefore
have driven great performance in moving brand metrics.
Due to the very personal nature of the mobile device, there are even higher
sensitivities around data collection and use for advertising in mobile than there
are in other channels. Different countries have widely varying stances on
mobile data when it comes to privacy so advertisers must ensure that their
partners are adhering to local privacy law.
2. Geo-targeting
Hyper-local information paves the way for better targeting and creative
optimisation, as marketers can now use beacon technology to target a
consumer as soon as they walk into a specific area, enabling advertisers to be
more accurate in terms of real-time and relevant communications. This is
particularly powerful at driving footfall, and results can be enhanced with
incentives such as vouchers or discounts. This has been talked about since
the early days of WAP phones but is now surprisingly easy to deliver and
effective.
It is also possible to work with data management platforms (DMP) who
maintain stock availability such that the geo-fences are dynamic and only
show ads in close proximity to outlets who are currently holding stock of
specific items.
Advertisers need to consider the regularity with which they use geo-targeting
however. Targeted messaging that produces hyper-relevant information will
engage consumers and build loyalty, but as soon as consumers see
messaging as spam, advertisers risk losing them for good. By combining
precise location with time and context, such as demographic information and
time of day, marketers can deliver highly relevant messages to the right
person at the right time and in the right place. Perhaps more importantly, they
can stop and start delivering them dynamically to reflect user behaviour and
changing context.

58
3. Native Advertising
The power of native advertising rests in the fact that ad units are designed to
blend with the surrounding environment seamlessly. By requiring just images
and text snippets - which can be combined and assembled programmatically
by a publisher on a per impression basis to integrate with the form and
function of their sites or apps - access to mobile advertising has become more
accessible.
Advertisers benefit from being able to use the standardised text and image
assets without having to create multiple different advertising units, and
publishers profit from the flexibility of ad assembly to find the best advertising
experience for their users and the most effective formats for their consumers.
One only has to look across social media platforms and any number of other
content environments to see how the same image and text assets can be
assembled into a multitude of different formats - recommendation widgets, in-
feed units, sponsored links and content tiles, for example.
With the push towards standardisation of the real-time bidding (RTB)
specifications for native advertising and large publishers offering self-serve
access into their native marketplaces – either directly through their own user
interfaces (UIs) or though API tool providers - native advertising is facilitating
ease of bookings across the mobile programmatic space and offering
advertisers immersive ways to engage consumers with increasing efficiency.
It is important to ensure that native advertising or sponsored content is clearly
identifiable from surrounding editorial content, as required by EU law in the E-
Commerce Directive.
4. Tracking Performance
Mobile has struggled in the past with perceptions that tracking is more difficult
or less effective, due to the absence of cookies. However, it is possible to
track activity on mobile that no other channel can track, including:
 App downloads / engagements: Typically, these are recorded by a
number of mobile tracking companies who are able to track activities
from most ad networks and demand side platforms (DSPs). This
means that advertisers need only integrate a single SDK into their app,
enabling the tracking companies to monitor all aspects of
measurement of performance.
 Web tracking: Tags from the big desktop ad servers are used in order
to track performance of display campaigns to drive performance on
websites. These can drive scale and great performance, particularly
when running ads on tablets.

59
 Real-world impact tracking: It is possible to track the impact of ads
on consumer behaviour in the real world. Whether this be determining
a ‘visit-rate’ to specific locations, or a stage further and understanding
whether purchase activity has taken place (which can rely on beacon
technology or coupon tracking).
Whilst these mobile specific tracking metrics are able to give insight into the
performance of a campaign within the mobile channel, marketers can gain a
more in-depth view into the consumer journey by using data platforms which
can evaluate interactions across a number of channels. In turn, this will allow
marketers to relate campaigns back to the bottom line and drive sales by
allocating the right budget to the right channel, and optimising campaigns in
real-time.
High-performing digital video campaigns are increasingly
key to building brands, especially as video consumption
continues its rapid shift from television to digital sources.
As advertisers try to find and engage their video audiences
wherever they happen to be watching: on their phone,
tablet, PC, or TV, programmatic technology becomes as
relevant here as it has been in desktop and mobile for
several years.
However, video advertising inventory is still in short supply and programmatic
technologies maturing. Here are some strategies to get the most from this
part of programmatic in video:
 Focus efforts on where the audience is spending their time
Seeding a video programmatically helps lay the groundwork for viral
sharing. Today, videos such as the popular Old Spice commercials are
often mistakenly assumed to have achieved viral status purely through
organic growth on YouTube and Facebook. In fact these campaigns
ran programmatically, using the distribution method to generate more
growth after seeing an initial positive organic reaction. Choose the
high-impact publishers to be associated with and secure guaranteed
placement on their site or app. Through a private marketplace (PMP),
marketers have pre-negotiated access to prominent video inventory
with top publishers—and their audiences—generally not available in
open auction environments.
4.3
Programmatic
Video

60
 Focus advertising on inventory that’s performing well
Programmatic video can be optimised in-flight – it provides flexibility for
media planners to highlight compelling content, and discover
unforeseen sectors of resonance for audience expansion. Marketer’s
should start with a list of their top inventory and refine it while the
campaign is running. By creating a white list of handpicked URLs or
mobile apps that are meeting the goals of the campaign, results can be
improved while reducing effective cost per mille (eCPM).
 Eliminate undesirable inventory
Develop a black list to prevent ads from appearing on certain URLs or
apps. For example, exclude competitors’ websites and apps,
publishers whose views don’t sync with the brand, or sites that don’t
meet the publishing or editorial standards—to protect the integrity of
the brand while improving the ROI of the campaign.
 Frequency controls
Every ad has an ideal frequency. Under-delivery will fail to build strong
brand awareness in the consumer's mind, while over-delivery can turn
those extra ad impressions into a negative brand association due to
fatigue and annoyance. Managing a video campaign through a real-
time media platform will enable advertisers to quickly learn the optimal
frequency for a campaign, but also deliver that optimal number through
RTB and the enforcement of universal frequency caps.
 Target campaigns by time of day and geography
Audiences move seamlessly from desktop to smartphone to tablet
throughout the day—and knowing when to reach them on which device
can be critical to the success of a campaign. A day-parting strategy
might, for example, target smartphone and tablet users with in-app
video ads during the evening hours and target desktop users during
daytime hours. Performance can be driven by removing low-performing
geographic regions from the target audience.
Integrated Campaigns: Video’s Effect on Other Channel Performance
While video alone is not a ‘lower funnel’ performance driver, it can still boost
the performance of other mid- and lower-funnel channels, such as display and
search. By monitoring performance lift on display and search through tracking
tags on an advertiser’s site, it is possible to optimise user targeting on video.
This means showing more video ads to the types of users who are known to
convert more often via search and display after video ad exposure. In RTB

61
language, this means bidding higher for specific users so that the ad message
remains top of mind for the consumers who are most likely to be influenced.
Audience-related insights gained from one channel -- display or search, for
example -- can shed new light on how to go about reaching the audience with
video. For example, an advertiser running a direct response campaign saw a
significant lift in conversions when it paired video and display. The advertiser
found that running one highly effective video cost less than the multiple
display and search ads that would have been required to achieve the same
results.
In 2014 the World Federation of Advertisers (WFA) Guide
to Programmatic included a ‘waterfall’ diagram showing
how advertiser’s budgets flow through the programmatic
ecosystem from their pockets to the pockets of the site
owners on which their ads appear.
Diagram 2 - WFA Money Flow Diagram9
9 WFA Guide to Programmatic Media, World Federation of Advertisers (2014) -
http://www.wfanet.org/media/programmatic.pdf
4.4
Understanding
the Value
Chain

62
Traditionally, a publisher would expect to see between 80-90% of the money
on any campaign, but as programmatic technology makes targeting more
efficient there is a shift in the value chain. The technology and services
discussed in this whitepaper are not without costs, with both buyers and
sellers investing in technology to enable them to operate programmatically.
The WFA report sparked an industry debate as many marketers were
unaware of the changes in value chain. Those discussions added some
useful context to the visual and are worth being understood by advertisers and
publishers before they finalise their programmatic approach.
Firstly it should be noted that the WFA money flow illustration is just that, a
hypothetical breakdown of how money can be allocated in a programmatic
campaign. It is missing the most important metric, which of course is the
campaign outcome. After all, the marketer’s job is not to fund the publisher but
rather to make the most effective use of their budget to drive the desired
outcome. If the ‘working’ media spend through programmatic is less than
through traditional buys, but because of better targeting and relevance the
advertising delivers greater results for the same or lower level of overall
spend, then the marketer is doing their job. The change in the spend mix is a
concern for the publisher, but for the advertiser it reflects the best use of
technology to drive campaign performance.
However, if the campaign results are not exceeding the traditional campaigns
then the marketer is not making effective use of the programmatic technology
and the money being diverted away from working media is effectively wasted.
From the publisher’s perspective, the potential halving of their income per
impression is not sustainable, especially when combined with their own costs
to deliver programmatically if using a supply side platform (SSP) (not shown in
the WFA illustration). In this paper we have talked about the adoption of
programmatic models and this is certainly one of the more influential factors
on publisher strategy – with data being used to enhance their value-add and
in-house programmatic teams to reduce costs and take direct control of their
programmatic direction.
Some stakeholders are intrinsic to programmatic trading and it can’t take
place without them. This doesn’t mean that there are not methods of reducing
costs across them, but a demand side platform (DSP) or an exchange / SSP
for example are typically needed if trading is to take place. There is also
compelling evidence that a ‘build your own’ strategy for the key transactional
technologies in programmatic will increase costs and reduce flexibility.

63
For the buyer, it is important to understand the value each stakeholder brings.
Every partner will argue that they bring value, but not all are essential to every
business or every individual campaign. For those that are, the cost of their
service cannot outweigh the value that they bring to the transaction. As noted
earlier in this white paper, these core technologies can make money in many
different ways and understanding this is a vital first step in programmatic.
There is immediately identifiable room to improve clarity and understanding in
trading relationships across the aptly named ‘value-adds’ partners that provide
verification, measurement, analytics, data, viewability, media planning,
attribution, and so on. It’s not to say that these innovations are unnecessary or
don’t add value, they do, but there is a relevance to particular trading
conditions that needs to be considered and can provide a cost efficiency if
addressed correctly.
Is the Service Required?
Taking a very basic example; a trusted premium publisher with a single ad unit
above the fold offering only O&O supply, with a long standing trading
relationship with the buyer. Is there really a requirement for viewability here, or
in fact many of the brand safety features that the buyer may be applying to all
their programmatic partners regardless of their individual trading situations?
The broad answer to this is no, but the question then becomes what method
does the tech and / or service provider have in place to target its offering
specifically where it’s needed as opposed to providing it in an unnecessary
blanket method? If the tech and / or service provider doesn’t have a scalable
solution and instead requires a more manual solution, or no solution at all,
then advertisers should not be paying for their shortcomings. This needs to be
negotiated out, and advertisers should look for a solution that does, or at the
very least get timings for when they will address it that they can be held to.
Double Counting
This should also be extended to services being paid for by both sides. It’s not
unheard of for the same service to be paid for twice at different points in the
sales / technology chain. For example brand safety is something taken
seriously by all stakeholders and there are benefits to all for using it. The buy-
side may use it to buy with confidence, protecting brand image and ensuring
placement on high quality engaging content. Meanwhile the sell-side might
use it to identify and prove to buyers its most premium opportunities or in
taking steps to greater transparency. However if both are using it
independently there is a high chance of both incurring a cost. A situation could
and does easily occur when a publisher already uses brand safety to ensure

64
its supply meets its standards, then likewise a buyer may insist on applying
the same or similar brand safety for a particular client or indeed all of its
buying because it also buys across publishers that don’t.
It can be far from an easy process to identify this, and then have the
conversation to negotiate how it will be resolved and implemented, but it’s
certainly a worthwhile one with regard to reducing margin.
Consolidate Services
Having identified which solutions add value and where they need to be applied
the next step should be to consolidate services where possible. There are
many advanced services in market that will argue their relevance and even
necessity in the ecosystem. As a specialist they may provide a service to a
higher level than others who may provide it as an add-on to an existing
broader tech. A broader system/service could offer financial economies of
scale that outstrip the value garnered from multiple specialists. Choosing a
vendor that offers brand safety, viewability, attribution and audience
verification will likely be cheaper than using multiple vendors with individual
specialisms. Unifying services in this way should provide a saving where
distinct services will likely inflate the cost, particularly if minimum revenue
commitments are in place which is where most cost savings are to be gained.
Payment Model
More attention to what fees are calculated against could also go some way to
alleviating what some have described as the ‘technology tax’. Typically fees
will be calculated as a percentage fee based on revenue but its good practice
to question if this is the most prudent way of doing things. On the surface this
seems fine, they make more money as the advertiser makes more money, but
the investment to this increase isn’t always balanced. Focusing on two
different sized ad placements can show a disparity in this investment which
supports the argument for paying on an impressions basis as opposed to a
percentage of revenue. A 468x60 format commands a lower CPM than larger
brand building formats such as the IAB Europe Brand Builder10
and IAB Rising
Star ad formats. An advertiser is paying a higher price for a larger piece of
advertising real estate, and a publisher is being compensated for giving away
10 IAB Europe PC and Tablet Brand Builder formats -
http://www.iabeurope.eu/files/3713/9530/9256/IAB_Europe_Brand_Builders_Ad_Formats_De
scriptions.pdf
IAB Europe Mobile Brand Builder formats -
http://www.iabeurope.eu/files/7114/3083/8067/IAB_Europe_Mobile_Brand_Builders_Ad_Form
at_Recommendation.pdf

65
prime editorial space. However, in most cases the cost of the ‘value-add’
service doesn’t increase based on the value of the ad placement so why
should a higher fee be taken? In some cases it may make more sense to pay
based on an impression basis than as a percentage of revenue, this could be
significant based on where the business is likely to grow.
Summary
In conclusion publishers can continue to add value through their
understanding and analysis of buying behaviours and their ability to deliver a
more tailored offering for advertisers as digital advertising evolves towards a
programmatic ecosystem.

66
The previous sections outlined some key considerations for developing a
programmatic strategy for buy-side and sell-side stakeholders. It is important
to also look at what strategies are currently adopted and how stakeholders
currently view programmatic advertising.
IAB Europe’s Attitudes towards Programmatic Advertising survey asked brand
advertisers, agencies and publishers about their adoption of and attitudes
towards programmatic advertising. The survey received responses from over
1,000 industry stakeholders and the full results will be published by the end of
July 2015.
The purpose of the following section is to provide an overview of buy-side and
sell-side stakeholders in ‘advanced’ programmatic markets. Advanced
markets are defined as those markets that are mature in terms of
programmatic adoption and includes: UK, Netherlands, France, Denmark,
Sweden, Norway, and Finland.
The majority of stakeholders in these markets are from companies of up to
100 employees and are responsible for budgets or revenues of between €1
and €10 million.
Current Adoption and Strategies
Almost three quarters of all stakeholders in the ‘advanced’ markets cite that
less than 40% of their digital advertising revenue is currently traded
programmatically. Whilst these markets are advanced in terms of
programmatic adoption, the survey findings suggest that there is still a lot of
potential to be fulfilled in terms of moving digital advertising trading to
programmatic mechanisms.
The sell-side are developing in-house strategies and are trading
programmatically in-house more than any other stakeholder in the advanced
markets; currently 61% of publishers state that more than 81% of their
programmatically traded display advertising revenue is in-house.
5 Current Adoption of Programmatic Advertising

67
On the buy-side then, the budget holders are more likely to outsource, for
example half of advertisers are outsourcing more than 81% of their
programmatically traded display advertising investment.
Indeed in-house is a strategic consideration for stakeholders; of those who
don’t already have an in-house strategy, 29% of advertisers are considering
one; 21% of agencies and 37% of publishers (see chart below). It is important
to note that 66% of agencies already have an in-house strategy suggesting
they are building up programmatic capabilities in these markets.
Chart 1: Stakeholders considering bringing programmatic in-house
There are a number of considerations for developing an in-house
programmatic strategy, such as budget and staffing. 60% of advertisers
considering an in-house strategy cite that they have sufficient budget to invest
in programmatic technologies and 40% believe they have enough budget for
the staffing requirements. 40% are unsure whether they have sufficient budget
for staffing suggesting they may need more education on the actual
requirements.
Agencies seem a bit more prepared with 88% of those considering in-house
citing they have the required budget for developing their programmatic offer,
and 65% say they have the required budget for staffing.
87% of publishers have the required budget to develop their programmatic
inventory strategy, however just under 50% have the required budget for
staffing and just over 40% are still unsure. Again, more education on the
37%
21%
29%
Publishers Agencies Advertisers

68
requirements for developing a programmatic strategy in terms of resources
may help these stakeholders to evaluate whether they can invest further.
The motivations for an in-house strategy of the buy and sell side are outlined
below.
Transparency of campaigns and control of data are important to advertisers
and the key reasons for these stakeholders to consider an in-house strategy.
By considering an in-house strategy, agencies want to increase their
operational control and be able to better access audience insights to enhance
their understanding of the consumer.
Publishers cite integration of audience data into their trading processes and
more efficient sales and ad operations processes as their key reasons for an
in-house strategy.
It is clear that by bringing programmatic trading in-house all stakeholders hope
to gain control of data and processes.
Attitudes towards Programmatic Advertising
Stakeholders in the markets defined as ‘advanced’ cite targeting efficiencies,
control and driving down costs or maximising media value as key reasons for
investing in programmatic. Both advertisers and agencies state targeting
efficiencies are their top reason, with advertisers citing lower cost of media as
their second and agencies increased granular control of media as their second
reason. It is clear that the general opinion of programmatic on the buy side is
that it can bring about efficiencies in both targeting and costs and control.
Publishers’ priorities are to maximise value of their media and secondly gain
trading efficiencies. Interestingly 55% of publishers in these markets want to
make their premium inventory available at scale to enable brand advertising
budgets to shift into digital channels.
Over three quarters of agencies and publishers in these markets place
programmatic in their top five or as their top priority.

69
Chart 2: Stakeholders that place programmatic in their top five or as
their top priority
Future of Programmatic Investment
The survey asked about barriers to adoption of programmatic and hiring and
training the people with the right skill set was the key barrier for all
stakeholders in the advanced markets. This again points to the need for more
education in the market. However despite this programmatic trading
investment and revenues are set to increase over the next 12 months and in
fact 31% of publishers see an increase of more than 51%; 28% of agencies
and 13% of advertisers also see an increase of more than 51%.
Summary
Having explored these advanced programmatic markets there is still great
potential for programmatic trading with the majority of digital advertising
investment still traded via manual processes. However the future looks
positive with all stakeholders citing an increase in trading via programmatic
technologies. The strategies for programmatic trading differ across the
stakeholder groups as the sell-side lean more towards in-house and the
budget holders outsourcing the majority. Agencies are also building up in-
house capabilities. An in-house strategy is seen as favourable by the buy-side
and sell-side stakeholders for control, transparency and efficiencies. As
46%
78%
85%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
Advertisers
Publishers
Agencies

70
outlined in this white paper, the business should assess which strategy is
most suited for their business and the decision tree in section 2.1 can be used
for this. Whilst the industry is positive about the future of programmatic
trading, there still remains some barriers to overcome, notably ensuring the
right skills are developed.
Please note – the full report exploring all stakeholders and less advanced
markets will be published later in July.

71
This white paper has explored the many facets of programmatic and shown
how it has the power to unlock audience and engagement, efficiency and
revenue.
IAB Europe’s final recommendations to keep in mind when travelling the
‘Road to Programmatic’ are detailed below:
 Start by thinking about audience and objectives so that programmatic
can act as a strategic tool
 Find out which programmatic operational model is best suited to the
company
 Think about how the data strategy fits – it is essential to have
complementary trading and data strategies
 Decide how to follow the consumer journey and how to feedback the
information from consumer to refine the strategy
 IAB Europe’s White Paper on Viewable Impressions11
sets the scene
for companies wanting to trade on viewability – watch out for more
developments in this space
 Consider using ad formats, e.g. IAB Europe’s Brand Builders12
, which
lend themselves well to programmatic methods and anticipate more
cross-media campaigns
11 IAB Europe Viewable Impressions White Paper (February, 2015) -
http://www.iabeurope.eu/news/iab-europe-publishes-viewable-impressions-white-paper
12 IAB Europe PC and Tablet Brand Builder formats -
http://www.iabeurope.eu/files/3713/9530/9256/IAB_Europe_Brand_Builders_Ad_Formats_De
scriptions.pdf
IAB Europe Mobile Brand Builder formats -
http://www.iabeurope.eu/files/7114/3083/8067/IAB_Europe_Mobile_Brand_Builders_Ad_Form
at_Recommendation.pdf
6 Conclusions and Recommendations

72
 With predictions anticipating nearly 1 billion smartphones in Europe by
202013
make sure that mobile is a key consideration when developing
a strategy
 As digital video increases in scale and takes a greater share of total
video consumption so the opportunity to engage with programmatic
video increases
 To increase learning ask the experts and engage with IAB Europe and
the local IABs across our network
13 Source: IHS Technology

73
IAB Europe would like to thank the authors from the following companies and
associations who took the time to contribute to this White Paper.
James Collier, Managing Director
EMEA, AdTruth
Marc Gosschalk, Partner Marketing
Manager, AdTruth
Graham Wylie, White Paper Leader and
Chairman, IAB Europe Programmatic
Trading Committee and Senior Director
EMEA and APAC Marketing, AppNexus
Joel Livesey, Director Business
Development and Privacy,
AudienceScience
Duncan Trigg, Vice President
Advertising Effectiveness, comScore
Alison Fennah, Executive Business
Advisor, IAB Europe
David Frew, Digital Programmes
Manager, IAB UK
Ben Crain, Chief Strategy Officer,
Improve Digital
Bastiaan Spaans, SVP Commercial,
Improve Digital
7 White Paper Contributors

74
Victoria Chappell, Marketing Director,
Integral Ad Science
Niall Hogan, UK Managing Director,
Integral Ad Science
Aimee-Marie Morgan, Senior Marketing
Manager EMEA, MediaMath
Emma Williams, Senior Director,
Regional Marketing EMEA, APAC &
LatAM, MediaMath
Oliver Gertz, Managing Director
Interaction EMEA and Programmatic
Lead Global Clients, MediaCom
Artur Banach, Programmatic Task Force
IAB Poland and CEO Netsprint
Augustin Decre, Managing Director
Southern Europe, nugg.ad
Emma Newman, Senior Director
Marketing International, PubMatic
Amit Kotecha, Head of Marketing EMEA,
Quantcast
Alex Merwin, VP Global Programmatic,
SpotXchange
Joanna Burton, Marketing Director
EMEA, SpotXchange
Marie Giesbert, Marketing Director,
StickyADs.tv

75
Gavin Stirrat, Chief Operating Officer,
StrikeAd
Nikki Hawke, Vice-President Marketing
Global, The Exchange Lab
Helen Miall, Head of Marketing EMEA,
Turn
Cadi Jones, Director of Audience
Specialists Northern Europe, Yahoo!
Mick Loizou Michail, Senior Manager
Mobile and Video Solutions, Yahoo!
Rahman Mohammed, Head of Strategic
Services, Yahoo!
Contact:
Marie-Clare Puffett, White Paper Project Manager – puffett@iabeurope.eu
Connect with us:
@IABEurope
IAB Europe
www.iabeurope.eu

76
Ad Exchanges*
Online, often highly automated auction-based marketplaces that facilitate the
buying and selling of inventory across multiple parties ranging from
advertisers, direct publishers, ad networks and Demand Side Platforms
(DSP).
Blacklist
A list of blocked items, terms or URLs to be used against the delivery of an
advertising campaign.
Data Management Platform (DMP)*
Platforms that allow advertisers, agencies, publishers and others to control
their own first-party audience and campaign data, compare it to third-party
audience data, and give the ability to make smarter media buying and
campaign planning decisions via behavioural targeting or extending audiences
via lookalike modeling. Advertisers and agencies generally utilise DMPs in
order to buy more effectively while publishers typically utilise DMPs in order to
segment their audiences and sell more effectively.
Demand Side Platform (DSP)
A technology platform with a bidding algorithm. Using a DSP, buyers
(advertisers, trading desks, agency trading desks) are able to centralise their
media buys with the programmatic purchase of digital inventory using a unified
platform across various SSPs, ad exchanges, and ad networks. A DSP is
designed to bid the optimal CPM for a particular impression in real-time,
incorporating the calculated value of the inventory against the campaign
goal(s). A DSP works in the interest of a buyer. In many cases a DSP offers
additional services such as centralized billing.
Look-alike Modelling
The process of finding users who resemble other users, for example a brand's
existing customers or converting users, based on their online behaviours and
known attributes.
Pixel
8 Glossary

77
A tracking or retargeting pixel is a small piece of code contained in a single
clear pixel on a website or specific ad unit that drops a cookie on the user's
browser. That cookie is used for tracking purposes.
Private Marketplace (PMP)
A unique set of biddable inventory (differentiated placements or data) that is
not available in its full transparency in a general RTB auction. PMP inventory
is packaged within a publisher's SSP and offered to selected buyers/bidders.
A buyer can use a deal ID (a universal identifier/numeric keycode) to gain
access to and bid on the specific PMP inventory.
Programmatic
An automated process for advertisers to buy media and an automated
process for media owners to sell ad inventory (including display, mobile and
video).
Real-Time Bidding (RTB)*
RTB is a protocol that enables the valuation and bidding on individual
impressions in real-time. The buying takes place over online media exchanges
– basically media marketplaces – which connect sellers (publishers) and
buyers (advertisers).
Supply Side Platform (SSP)
A technology platform that makes it possible to sell automated online media to
different parties. Using an automated yield optimiser, the algorithms in an SSP
ensure that the publisher receives the highest turnover per impression. An
SSP is connected to multiple demand sources which include DSPs, ad
exchanges and advertising networks. An SSP always works in the interest of a
publisher. In many cases an SSP also offers additional services such as
commercial agreements with ad exchanges and centralised billing.
Site Extension
The process of looking at the pages of a site that the audience has been active
on, to gain understanding of site visitors so that they can be retargeted more
effectively.
Trading Desk*14
An agency branch trading entity known as the expert operators in their use of
new technology. These entities can be independent or operate within an
agency holding company. This group of people (known as traders) play the
14 Please note – terms marked with a * are from the IAB UK Jargon Buster -
http://www.iabuk.net/resources/jargon-buster

78
day-to-day campaign management role. Who uses: Agency holding
companies, operating agencies, advertisers.
Whitelist
A list of approved items, terms or URLs to be used against the delivery of an
advertising campaign.

The Egg
Bara Street 175
1070 Brussels, Belgium
+32 (0) 2 526 55 68
White Paper enquiries: puffett@iabeurope.eu
General information: communication@iabeurope.eu
Twitter: @IABEurope LinkedIn: IAB Europe www.iabeurope.eu
About IAB Europe
IAB Europe is the voice of digital business and the leading
European-level industry association for the digital advertising
ecosystem. Its mission is to promote the development of this highly
innovative sector by shaping the regulatory environment, investing
in research and education, and developing and facilitating
the uptake of business standards. Together with its members
– companies and national trade associations – IAB Europe
represents over 5,500 organisations.
europe

Iab europe road_to_programmatic_white_paper_july_2015_15.07.15

More Related Content

What's hot (19)

Viewers also liked (12)

Similar to Iab europe road_to_programmatic_white_paper_july_2015_15.07.15 (20)

More from IAB Europe (20)

Recently uploaded (20)

Iab europe road_to_programmatic_white_paper_july_2015_15.07.15