Idashpage.

com

What works in Internet Advertising

*A Limited Edition Report

:::avant |marketer:::

:::avant |marketer:::

INTRODUCTION
"What works in Internet Advertising?" Ask this simple question, and you're likely to be hit by a barrage of emotionally charged answers. In many cases, you even get hit by questions. Questions that question the very question at hand: "Does Internet Advertising even work at all?" they ask. In connection with no other medium has there been more detailed study done into the effectiveness of advertising, and what works and what doesn't, than there has been in connection with the Internet. And, yet, in no other medium have the controversies over what works (and whether the medium itself works for advertising at all) remained heated so long, virtually unchanged. There seems to be a glaring discrepancy: So many facts. So little understanding. While it's no doubt true that those of you who deal in Internet Advertising daily have a much clearer answer to these questions than many of your traditional peers who have been slow to experiment online, and certainly a clearer idea than many in the press, there is still a tendency on all sides to make strategic decisions based on opinion. The likely reason for this is somewhat ironic: There is so much information being put out these days on Internet Advertising effectiveness, that we're being drowned in it. There is simply such a large amount of research coming out that few have the time to give it any thought, let alone to use it to our strategic benefit. And yet, ultimately, it still remains true that following the research can improve your bottom-line advertising ROI. In other words, a lot of this research is truly valuable. But, how do you sift through it to find the gold? To do our small part to remedy this situation, in what follows in this limited edition report, we arm you, the reader, with a few critical, choice facts and some strategic thinking on them that address the question of what works in Internet Advertising today. Certainly, there is more to be said on what works than what is printed in this brief report. But, what are included here are at least some clear-cut, usable answers on what is effective, and more importantly cost-effective at a fundamental level, in Internet Advertising, right now.

2002 avant|marketer. All Rights Reserved.

Page 1

:::avant |marketer:::

To be more specific, what we have done in this report is reviewed what we believe are three of the most important pieces of Internet Advertising effectiveness research that have been released in the recent past. We have used these to bring you actionable conclusions on three important questions: 1. Does the Internet work for branding? And, how exactly does the Internet compare to the traditional media (Television and Print) for branding, both in terms of effectiveness and costeffectiveness? 2. Which Internet ad formats are the most effective? Which are the most cost-effective? And, are larger ad formats really worth the cost? 3. How can Internet ad creative be tangibly adjusted, to significantly improve campaign ROI? Our review of each research piece is followed by our own brief analysis, which is based, in part on research done here at avant|marketer, and, in part on our numerous of hours worth of exclusive interviews with agency heads, Internet Advertising industry executives, researchers, and the like. Also, to make this report even more useful, we have added references to similar important research after our review of each of the three pieces. In all cases we have included "immediate download" links to the research mentioned, right in this report - which puts an additional set of critical data and facts at your fingertips, such that this report could even be looked at as a roadmap to much of the most important Internet Advertising effectiveness research that has been done, todate. Some of the conclusions that you find below may be somewhat obvious. Others will likely surprise you. But, whatever the case, you will be armed with the facts, and will be empowered to act accordingly. To your success, Ajay Segal (Editor, avant|marketer)

2002 avant|marketer. All Rights Reserved.

Page 2

:::avant |marketer:::

SECTION 1: Does Internet Advertising Work?

Research Review:
Correction: Yes, Internet Advertising works, but... Morgan Stanley Dean Witter February 2001

OVERVIEW
This is the Morgan Stanley Dean Witter Internet Advertising effectiveness report that has been very widely discussed since its release (though far less widely read than it has been discussed - which is unfortunate). Without a doubt, this report, demands to be read through by every brand manager, media buyer and planner, and executive dealing in Internet media - if not for any other reasons than the following: 1) It succinctly and completely answers some of today's most critical questions surrounding the effectiveness of Internet Advertising: - Does Internet Advertising work? - How powerful is Internet Advertising for branding? - How powerful is Internet Advertising as a direct response tool? - How does the Internet stack-up against the traditional media, such as Television? 2) It represents a comprehensive analysis of the findings of most of the major Internet Advertising effectiveness studies through February 2001, from sources such as Dynamic Logic, Ad Relevance, 24/7 Media, DoubleClick, Morgan Stanley Dean Witter, the Direct Marketing Association, and Media Dynamics. Therefore, the findings of this report represent the cumulative work of multiple independent research organizations, over a very substantial total sample population.

2002 avant|marketer. All Rights Reserved.

Page 3

:::avant |marketer:::

KEY FINDINGS
Boiled down to the most critical essentials, the findings of this report are that:

+ Banner Advertising is a highly effective branding tool.

- Banner Advertising is more effective than Magazines, Newspapers, and Television at increasing Brand Recall.

Effectiveness (Brand Recall) Medium Banners Magazines Newspapers Television Lift (first exposure) 27% 26% 23% 17%
Source: Morgan Stanley Dean Witter

- Banner Advertising ranks roughly in-between Television and Magazines increasing Product Interest.

Effectiveness (Product Interest) Medium Television Banners Magazines Lift (first exposure) 46% 44% 44%
Source: Morgan Stanley Dean Witter

2002 avant|marketer. All Rights Reserved.

Page 4

:::avant |marketer:::

- Banner Advertising is not as effective as Television and Magazines for increasing Brand Awareness.

Effectiveness (Brand Awareness) Medium Television Magazines Banners Lift (first exposure) 36% 29% 14%
Source: Morgan Stanley Dean Witter

+ When calculated using current rate card rates, Banner Advertising is less cost-effective than
Television, Magazines, and Newspapers for increasing Brand Recall, Product Interest, and Brand Awareness. However, and most importantly, when considered against the actual going market rate for Banners, Banner Advertising proves to be 40-80% more cost-effective than Magazines and Television for increasing both Brand Recall and Product Interest, although still less cost-effective than Magazines for increasing Brand Awareness (by about 20%).

Relative Cost-Effectiveness for Branding Branding Objective Increase Brand Recall Cost-Effectiveness Ranking (by Medium) 1. Banners (based on market rate CPM) 2. Magazines 3. Newspapers 4. Television 1. Banners (based on market rate CPM) 2. Magazines 3. Television 1. Magazines 2. Banners (based on market rate CPM) 3. Television
Source: Morgan Stanley Dean Witter

Increase Product Interest

Increase Brand Awareness

2002 avant|marketer. All Rights Reserved.

Page 5

:::avant |marketer:::

+ Direct Response Marketing online: In cases in which the goal of direct response marketing is taken to be sales (as opposed to subscriptions, information requests, etc.), Banner Advertising is only a moderately cost-effective form of direct response marketing, with evidence indicating that Banner clickthrough rates do not correlate well with sales rates.
However - though less cost-effective than Shared Direct Mail, various forms of Coupon Marketing, and Direct Email Marketing - Banner Advertising is still more cost-effective (in terms of cost-persale), than Solo Direct Mail.

+ Direct Email Marketing is a highly cost-efficient form of direct response marketing, measured in
terms of cost-per-sale. However, it must be recognized that this cost-efficiency is due partially to the fact that Direct Email Marketing is used frequently for customer retention, as opposed to customer acquisition purposes.

Relative Cost-Effectiveness for Direct Response Metric Cost-Per-Sale Cost-Effectiveness Ranking (by Medium) 1. Shared Direct Mail 2. Free Standing Insert Coupon 3. Direct Email Marketing 4. Point of Sale-targeted Coupon 5. Banners 6. Direct Mail
Source: Morgan Stanley Dean Witter

CRITICAL RAW DATA CONTAINED IN THIS REPORT

Over-and-above what has been mentioned here, be aware of that the following critical raw data is contained in this report:

+ Page 5: Comprehensive chart mapping the comparative effectiveness of Internet, Television,

Newspaper, Magazine, and a wide variety of forms of Direct Mail marketing/advertising (Coupons, Shared Direct Mail, and Solo Direct Mail), across a broad range of critical branding and direct response metrics. The value of this chart alone justifies having this report.

2002 avant|marketer. All Rights Reserved.

Page 6

:::avant |marketer:::

THE avant|marketer TAKE

It shouldn't be news that Internet Advertising works to some degree or another. This report establishes the precise degree to which it does work. And the fact that it does so by pulling data from a wide variety of sources, lends significant strength to the conclusions reached. avant|marketer believes that the single most mission-critical finding in this report is that not only does Banner Advertising work for branding, Banner Advertising is currently one of the most costefficient means of increasing Brand Recall and Product Interest, and is also highly cost-efficient as a means of generating Brand Awareness - more so in all cases than Television (as unintuitive as this might sound). Therefore, we conclude that advertisers, particularly advertisers with established brands seeking to reinforce those brands, that are not currently leveraging Banner Advertising to a degree proportionate to the total media time spent online by their target audience, are not achieving the overall advertising ROI that they could be, are leaving money on the table, and are at a significant competitive disadvantage. If anything, we believe, the conclusions of this report on the overall cost-effectiveness of Banner Advertising are even truer today than they were when Morgan Stanley Dean Witter originally published its findings in February 2001. Since that time, actual market CPM rates for Banners have fallen, though it is very unlikely that the branding capabilities of Banners have fallen proportionately. Therefore, we conclude, Banners are an even more cost-effective means of branding today, than they have been at any time, certainly in the recent past, and it's likely that advertisers currently neglecting Banner Advertising as part of their overall media mix are suffering to an extent even greater than was previously the case. Certainly, a brief note on Direct Email Marketing is also in order: Though this report concludes that Direct Email Marketing is, all uses considered, a less cost-effective sales generation tool than Shared Direct Mail and Free Standing Coupons, if Direct Email Marketing were looked at purely from a customer retention standpoint, we believe that Direct Email Marketing would prove to be the most cost-efficient means of Direct Response Marketing (for sales) currently available in any medium. Therefore, advertisers not already doing so must move to incorporate retention-oriented Direct Email Marketing into their overall marketing strategy (through the development of an opt-in email house file).

DOWNLOAD
Click here to begin immediate download of the full version of this report (14 pages; 115 KB; ).

2002 avant|marketer. All Rights Reserved.

Page 7

:::avant |marketer:::

IMPORTANT RELATED RESEARCH

For more analysis and raw data on issues related to the above, we recommend you access the following reports, which are available for immediate download:

+ 1997 IAB Internet Advertising Effectiveness Study

IAB / Millward Brown Interactive Click here to begin immediate download of this report (88 pages; 3871 KB; ).

+ Cross Media Research: Optimizing the Media Mix

MSN / IAB / ARF Click here to begin immediate download of this report (2 pages; 550 KB; ).

+ The Australian Online Advertising Effectiveness Study 2001

ACNielsen.consult Click here to begin immediate download of this report (29 pages; 604 KB; ).

2002 avant|marketer. All Rights Reserved.

Page 8

:::avant |marketer:::

SECTION 2: Which ad formats work?

Research Review:
Online Builds Brands IAB / Dynamic Logic July 2001

OVERVIEW
This critical piece of research provides important data on a perennial issue: Does size matter? In other words: How well do the larger ad formats really work? More specifically, this research tackles two critical questions for people dealing in interactive media: 1) Does the size of an Internet-based ad impact that ad's overall effectiveness for branding? 2) If so, to what extent? Importantly, the results of this research are based on a sample population of a relatively substantial size.

KEY FINDINGS
This report looks at the comparative effectiveness of the most important IAB-approved Internet ad formats, from small to large (468x60 pixel Banner units, 120x600 and 160x600 pixel Skyscraper units, and 240x400, 336x280, 360x300 and 300x250 pixel Large Rectangle units), across a host of critical branding metrics (Brand Awareness, Brand Favorability, Message Association, and Purchase Intent). In short, this report concludes that, in general, ad format size does impact the branding effectiveness of Internet ads across the majority of branding metrics. In essence: Larger ad formats (Skyscrapers and Large Rectangles) outperform the smaller units (468x60 Banners), in most contexts (there are exceptions).

2002 avant|marketer. All Rights Reserved.

Page 9

:::avant |marketer:::

The most important bottom-line findings of the study that we believe you should be aware of are these:

+ Skyscraper and Large Rectangle units are, on the whole, 3 to 6 times more effective than
standard Banner ads in creating a lift in Brand Awareness and Message Association.

Effectiveness (Brand Awareness) Format Skyscraper Large Rectangle Banner Lift (first exposure) 9% 9% 5%
Source: IAB / Dynamic Logic

Effectiveness (Message Association) Format Large Rectangle Skyscraper Banner Lift (first exposure) 73% 60% 19%
Source: IAB / Dynamic Logic

+ In the case of Brand Favorability, however - and this is a key exception - the findings show that the exact opposite is the case, and that Banners, in fact, outperform the larger creative units (both Skyscrapers and Large Rectangles) in terms of creating lift.

Effectiveness (Brand Favorability) Format Banner Large Rectangle Skyscraper Lift (first exposure) 13% 10% 8%
Source: IAB / Dynamic Logic

2002 avant|marketer. All Rights Reserved.

Page 10

:::avant |marketer:::

+ In the case of lifting Purchase Intent, findings show again that both Skyscrapers and Large Rectangle units outperform Banners.

Effectiveness (Purchase Intent) Format Large Rectangle Skyscraper Banner Lift (first exposure) 16% 14% 6%
Source: IAB / Dynamic Logic

CRITICAL RAW DATA CONTAINED IN THIS REPORT

Over-and-above what has been mentioned here, be aware of that the following critical raw data is contained in this report:

+ Page 2: Comprehensive chart mapping the effectiveness of three different types of Internet ad
format (Banners, Skyscrapers, and Large Rectangles). In addition to the lift figures provided above, this chart also displays the "percentage point" impacts caused by each of the three ad formats on this same range of branding metrics.

THE avant|marketer TAKE

avant|marketer believes that the conclusions of this report concerning the relative effectiveness of larger Internet ad formats in lifting Brand Awareness and Message Association are significant, and should be taken seriously by advertisers. We find it noteworthy that important research done independently of this report supports its conclusions regarding the impact of ad format on Brand Awareness (for instance, see the DoubleClick study; download link included below). However, that having been said, the question of advertising effectiveness should never be considered by advertisers in isolation from the question of cost. In other words: The real question, we believe, must be: Are advertisers justified in paying the, often, higher premiums associated with these larger units? Are these larger units cost-effective?

2002 avant|marketer. All Rights Reserved.

Page 11

:::avant |marketer:::

Building on the cost-effectiveness conclusions of the Morgan Stanley Dean Witter report (see above for summary), based on the data provided in the course of the Online Builds Brands report, with which we're presently concerned, and in light of current market CPM rates for Skyscrapers, avant|marketer believes that Skyscrapers represent a particularly excellent value for building Brand Awareness, and are more cost-effective for this purpose than are Banners, Television, Newspapers, and - most probably - Magazines (which Morgan Stanley concludes is the most costefficient means presently available of lifting Brand Awareness). Advertisers seeking to lift Brand Awareness must therefore, we believe, give serious consideration to the use of Skyscraper units. Finally, regarding the conclusions of this report on the impacts of ad format on Brand Favorability and Purchase Intent, we urge that advertisers move cautiously. Strangely, our assessment here is largely based on our review of the broader study released by the IAB and Dynamic Logic, themselves entitled Ad Unit Effectiveness Study (see below for download link), which finds that Brand Favorability and Purchase Intent are only significantly lifted after multiple ad exposures (3 to 4 exposures or more). While this does not necessarily call into question the relative effectiveness of the various ad formats (Banners, Skyscrapers, and Large Rectangles) in lifting Brand Favorability and Purchase Intent, it does call into question using any of these ad units to cost-effectively lift Brand Favorability and Purchase Intent (as, to put it plainly, more exposures cost more to deliver). Therefore, we urge advertisers to look carefully at the overall value (lift/dollar) structure of any Internet media buy aimed at lifting brand Favorability and Purchase Intent, rather than taking the results of this study on this, at face value.

DOWNLOAD
Click here to begin immediate download of the full version of this report (4 pages; 842 KB; ).

IMPORTANT RELATED RESEARCH

For more analysis and raw data on issues related to the above, we recommend you access the following reports, which are available for immediate download:

+ Online Marketing Effectiveness

DoubleClick Click here to begin immediate download of this report (3 pages; 205 KB; ).

2002 avant|marketer. All Rights Reserved.

Page 12

:::avant |marketer:::

SECTION 3: What creative strategies work?

Research Review:
The Role of Creative Execution in Online Advertising Success IAB / Rex Briggs October 2001

OVERVIEW
As the previous study proves, there are noticeable performance gains that accrue from the use of larger ad formats. However, there is also the separate question of, whatever format is being used, how to drive optimal performance (and, thereby ROI) from that format. One question that comes under this head is the question of ad creative. In other words: Are there concrete, proven guidelines that can be employed to adjust Internet ad creative for optimal branding performance? This report, written by Internet Advertising expert, Rex Briggs, pools findings on Internet ad creative from a variety of substantial studies done by the likes of Dynamic Logic, AdRelevance, 24/7 Media, and ACNielsen.consult, to answer this question with concrete guidelines.

KEY FINDINGS
According this research, in order to improve the branding performance of Internet-based ads, advertisers should adhere to the following set of actionable guidelines in constructing their ad creative:

+ Brand the first frame: Use their brand name, logo, and product picture on the first frame of an ad. + Brand at the top on Skyscrapers: In cases in which skyscraper units are being used, to
maximize branding impact from such units, advertisers should consider putting branding information (brand names, logos, and product pictures) at the top of the ad.

2002 avant|marketer. All Rights Reserved.

Page 13

:::avant |marketer:::

+ Ditch "click here": Eliminate the use of the phrase "click here" in ads, as the evidence indicates that this signals to consumers to ignore Internet-based ads. + Use light backgrounds: Consider constructing ads that employ a light-colored background. + Employ a high background/foreground contrast: Make certain that there is a high contrast between the background color of the ad, and its foreground color. + Avoid blinking logos: The evidence suggests that while, on the whole, dynamic (i.e. animated)
messages are okay, blinking logos might degrade the ability of Internet-based ads to effectively brand.

+ Use no more than 15 ad elements: Ads should be designed so as to not contain more than 15
distinct ad elements (i.e. discreet graphical or textual components).

+ Use large logos: The corporate or product logo (whichever is to be branded) should be
designed so as to take up at least 14% of the space within the creative unit it is being placed in.

+ Include a human face: Firms should incorporate an image of a human face into their ads. Evidence suggests that this impacts the branding capability of ads, particularly with regards to creating product/brand interest.

CRITICAL RAW DATA CONTAINED IN THIS REPORT

N/A

THE avant|marketer TAKE

While we have little to add to the research presented in this report, other than to "second" what has been said here, we would like to briefly mention a closely related area for those buying and creating Internet ads to consider, not mentioned here.

2002 avant|marketer. All Rights Reserved.

Page 14

:::avant |marketer:::

While we certainly believe that sufficient emphasis hasn't been placed on the importance of developing quality ad creative in line with proven best practices, a matter that has been even less considered than the role of creative design in driving branding ROI, is the impact of site design and page clutter on the ability of an ad to effectively brand. This is, what we'd like to call "beyond the banner clutter". To put it simply, there is evidence now mounting that placing an ad in a cluttered online environment significantly impairs the branding capability of that ad. In the realm of the print and broadcast media, media clutter has always been closely monitored, and its effects on ad effectiveness have been, too. In the online space, these factors have not been monitored with equal care. According to early data from a few sources, most notably data gathered in a recent Dynamic Logic study on page clutter (download link included below) and ad effectiveness data from Terra Lycos from their recent re-design and re-launch of the Lycos portal and network of sites, which sought to reduce page clutter (for details see: http://www.terralycos.com/press/pr_11_5e_01.html), there are two main sources of page clutter. First, page clutter can be caused when there are too many ads being served per page. And second, can be caused when pages on which ads are being served contain too many elements (discreet components, including colors). Although the evidence is not yet complete, avant|marketer believes that companies can obtain a significant competitive advantage by actively seeking out, and running their ads on sites that are low clutter, in terms of these two elements. This is the case as current pricing does not seem, on the whole, to reflect ad clutter (top ad slots on pages being sold to the same audience in terms of size and demographics appear to have roughly the same price attached, regardless of ad clutter), while it is likely that the low clutter pages provide a distinctly better return on branding investment (ROI) for the advertiser.

DOWNLOAD
Click here to begin immediate download of the full version of this report (7 pages; 417 KB; ).

2002 avant|marketer. All Rights Reserved.

Page 15

:::avant |marketer:::

IMPORTANT RELATED RESEARCH

For more analysis and raw data on issues related to the above, we recommend you access the following reports, which are available for immediate download:

+ The Five Golden Rules of Online Branding

Dynamic Logic / AdRelevance / 24/7 Media Click here to begin immediate download of this report (14 pages; 119 KB; ).

+ Visual Noise: The Role of Site Clutter in Advertising Effectiveness

Dynamic Logic Click here to begin immediate download of this report (24 pages; 186 KB; ).

2002 avant|marketer. All Rights Reserved.

Page 16

:::avant |marketer:::
GET YOUR FREE SUBSCRIPTION NOW!
avant|marketer is the first email newsletter devoted to high-level Internet Advertising strategy. Each week, readers count on avant|marketer to significantly enhance their knowledge of the breaking trends, ideas, technologies, companies, and people that are shaping the future of Internet Advertising, and to provide them with a high-level strategic understanding of what questions to think about and why, and how best to think about them. avant|marketer enhances our readers' understanding of the "big picture" of Internet Advertising, and shows them how they can successfully navigate the current industry turbulence, and win. Current readers of avant|marketer include top-level insiders at the world's most important agencies and companies, including: - DoubleClick - Saatchi & Saatchi - Proctor & Gamble - Circle.com - Avenue A

CLICK HERE TO SUBSCRIBE NOW

Idashpage.com