These materials are © 2020 John Wiley & Sons, Inc.

Any dissemination, distribution, or unauthorized use is strictly prohibited.

 Data
Sharing
2nd Snowflake Special Edition

by Lawrence C. Miller
and David Baum

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Data Sharing For Dummies®, 2nd Snowflake Special Edition

Published by
John Wiley & Sons, Inc.
111 River St.
Hoboken, NJ 07030-5774
www.wiley.com

Copyright © 2020 by John Wiley & Sons, Inc., Hoboken, New Jersey

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any
means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted under Sections
107 or 108 of the 1976 United States Copyright Act, without the prior written permission of the Publisher. Requests to
the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River
Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permissions.

Trademarks: Wiley, For Dummies, the Dummies Man logo, Dummies.com, and related trade dress are trademarks
or registered trademarks of John Wiley & Sons, Inc. and/or its affiliates in the United States and other countries,
and may not be used without written permission. Snowflake and the Snowflake logo are trademarks or registered
trademarks of Snowflake Inc. All other trademarks are the property of their respective owners. John Wiley & Sons,
Inc., is not associated with any product or vendor mentioned in this book.

LIMIT OF LIABILITY/DISCLAIMER OF WARRANTY: THE PUBLISHER AND THE AUTHOR MAKE NO REPRESENTATIONS
OR WARRANTIES WITH RESPECT TO THE ACCURACY OR COMPLETENESS OF THE CONTENTS OF THIS WORK AND
SPECIFICALLY DISCLAIM ALL WARRANTIES, INCLUDING WITHOUT LIMITATION WARRANTIES OF FITNESS FOR A
PARTICULAR PURPOSE. NO WARRANTY MAY BE CREATED OR EXTENDED BY SALES OR PROMOTIONAL MATERIALS. THE
ADVICE AND STRATEGIES CONTAINED HEREIN MAY NOT BE SUITABLE FOR EVERY SITUATION. THIS WORK IS SOLD
WITH THE UNDERSTANDING THAT THE PUBLISHER IS NOT ENGAGED IN RENDERING LEGAL, ACCOUNTING, OR
OTHER PROFESSIONAL SERVICES. IF PROFESSIONAL ASSISTANCE IS REQUIRED, THE SERVICES OF A COMPETENT
PROFESSIONAL PERSON SHOULD BE SOUGHT. NEITHER THE PUBLISHER NOR THE AUTHOR SHALL BE LIABLE FOR
DAMAGES ARISING HEREFROM. THE FACT THAT AN ORGANIZATION OR WEBSITE IS REFERRED TO IN THIS WORK AS
A CITATION AND/OR A POTENTIAL SOURCE OF FURTHER INFORMATION DOES NOT MEAN THAT THE AUTHOR OR THE
PUBLISHER ENDORSES THE INFORMATION THE ORGANIZATION OR WEBSITE MAY PROVIDE OR RECOMMENDATIONS
IT MAY MAKE. FURTHER, READERS SHOULD BE AWARE THAT INTERNET WEBSITES LISTED IN THIS WORK MAY HAVE
CHANGED OR DISAPPEARED BETWEEN WHEN THIS WORK WAS WRITTEN AND WHEN IT IS READ.

For general information on our other products and services, or how to create a custom For Dummies book for your
business or organization, please contact our Business Development Department in the U.S. at 877-409-4177,
contact [email protected], or visit www.wiley.com/go/custompub. For information about licensing the
For Dummies brand for products or services, contact BrandedRights&[email protected].

ISBN 978-1-119-66641-7 (pbk); ISBN 978-1-119-66645-5 (ebk)

Manufactured in the United States of America

10 9 8 7 6 5 4 3 2 1

Publisher’s Acknowledgments

We’re proud of this book and of the people who worked on it. For details on how to
create a custom For Dummies book for your business or organization, contact info@
dummies.biz or visit www.wiley.com/go/custompub. For details on licensing the
For Dummies brand for products or services, contact BrandedRights&Licenses@
Wiley.com.
Some of the people who helped bring this book to market include the following:
Development Editor: Nicole Sholly Production Editor: Siddique Shaik
Project Editor: Martin V. Minner Snowflake Contributors Team:
Vincent Morello, Daniel Kuperman,
Executive Editor: Steve Hayes
Leslie Steere, Michael Nixon,
Editorial Manager: Rev Mengle Matthew Glickman
Business Development Representative:
Karen Hattan

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Table of Contents
INTRODUCTION................................................................................................ 1
About This Book.................................................................................... 1
Icons Used in This Book........................................................................ 2
Beyond the Book................................................................................... 2

CHAPTER 1: Getting Up to Speed on Data Sharing Basics.......... 3

What Is Data Sharing?........................................................................... 4
Data Sharing Examples......................................................................... 7
How Organizations Share Data........................................................... 7
Exploring Data Sharing Possibilities.................................................... 8

CHAPTER 2: Understanding Traditional Data

Sharing Challenges...................................................................... 9
Addressing a Multi-Faceted Problem.................................................. 9
Dealing with Conventional Data Sharing:
Time-to-Value Delays.......................................................................... 12
Computing Complexity Challenges................................................... 14
Sharing Data the Conventional Way: Business Pain Points........... 15
Data sharing within an organization........................................... 15
Data sharing business-to-business (B2B)................................... 16
Monetizing data............................................................................. 16

CHAPTER 3: Recognizing the Business Value

of Sharing Data............................................................................. 17
Looking Back at the Early Days of Data Sharing.............................. 17
Assessing the Business Value of Data Sharing
for Organizations................................................................................. 18
Sharing data across LOBs............................................................. 18
Sharing between organizations: Outbound............................... 19
Sharing between organizations: Inbound................................... 19
Monetizing data............................................................................. 21

Table of Contents iii

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 CHAPTER 4: Enabling Live Data Sharing................................................. 23
Exploring Modern Data Sharing........................................................ 23
Winning with modern data sharing............................................. 24
Making data sharing easy............................................................. 25
Using Modern Cloud Data Sharing.................................................... 29
Learning How It All Works.................................................................. 29
Controlling Access to Shared Data by Using Secure Views............ 30

CHAPTER 5: Assessing the Impact of Modern

Data Sharing................................................................................... 33
Modern Approaches to Data Collaboration..................................... 33
The Democratization of Computing.................................................. 34
Data-Driven Business Decisions........................................................ 35
The Commercialization of Data......................................................... 36

CHAPTER 6: Using Exchanges to Share Data

and Monetize Insights............................................................ 39
Examining the History of Data Marketplaces and Exchanges....... 39
Moving into the Modern Age with Real-Time Data Sharing........... 41
Establishing the Right Architecture for Exchanging Data............... 42

CHAPTER 7: Monetizing Your Data............................................................. 45

Examining Industry Opportunities.................................................... 46
Getting in Front of Data Monetization Trends................................. 47
Observing Good Data Governance................................................... 48
Evaluating Distribution Channels...................................................... 48
Pricing Your Data Sharing Service..................................................... 49

CHAPTER 8: Governing Your Data............................................................... 51

Sharing the Same Data with One or Multiple Consumers............. 52
Sharing Different Subsets of Data with Multiple Consumers........ 53
Using Secure Views to Provide Predefined Slices of Data.............. 54
Answering Questions While Maintaining Data Privacy................... 55
Using Secure Joins to Combine Data between Companies............ 56

CHAPTER 9: Six Steps to Advance Your Business

with Modern Data Sharing.................................................. 57

iv Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Introduction
O
rganizations inside an enterprise acquire crucial insight
by analyzing data they share with each other. For exam-
ple, finance teams need sales data to forecast future
financial performance. Product management teams require mar-
keting data to determine future products and services. Executive
management needs up-to-the minute dashboards, fueled by data
from many parts of the enterprise, to make timely, data-driven
business decisions.

Outside an organization, retailers share sales data with their

vendors to manage inventory and supply chains. Software-as-a-
service (SaaS) providers share the data they collect with their cus-
tomers to provide them with deeper insights into their business.
Healthcare providers securely share patient data with vendors
that provide ancillary products and with other business partners
that analyze that data to help improve patient services. The list
goes on. According to Forrester Research, more than half of global
data and analytics decision makers report their firm is expanding
its ability to source external data.

Data has become more than something to collect and analyze.

It’s an asset you can easily and securely make available inside
and outside your organization to streamline operations, swiftly
deliver more-personalized customer experiences, and open
up new market opportunities. As a data provider, you can also
securely monetize your data and create self-service relationships
between your organization and an endless number of data con-
sumers. In fact, 47 percent of global data and analytics decision
makers report their organizations currently commercialize data,
according to Forrester.

About This Book

Welcome to Data Sharing For Dummies, 2nd Snowflake Special
­Edition, where you explore how modern data sharing enables
any organization to share and receive live data, within min-
utes, in a governed and secure way  — with almost none of the
risk, cost, headache, and delay that have plagued traditional data

Introduction 1

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 sharing methods. Modern data sharing allows an organization to
­easily and quickly forge one-to-one, one-to-many, and many-
to-many relationships to share data in new and imaginative ways
and reduce time to insight to a level never before possible.

Icons Used in This Book

In this book, you’ll occasionally see special icons calling attention
to important information. Here’s what to expect:

The case studies provide best practices from organizations that

have successfully used modern data sharing methods.

This icon points out information you should commit to your non-
volatile memory — your gray matter.

This icon explains the jargon beneath the jargon.

This icon points out useful nuggets of information and helpful

advice.

These alerts offer practical advice to help you avoid potentially

costly or frustrating mistakes.

Beyond the Book

At the end of this book, if you’re thinking, “Where can I learn
more?” just go to www.snowflake.com to find out what Snowflake
offers, obtain details about modern data sharing, view webinars,
get the scoop on upcoming events, and access documentation and
other support. You can contact Snowflake or even try its technol-
ogy for free.

2 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Defining data sharing

»» Recognizing the importance of

data sharing

»» Exploring data sharing examples

»» Understanding how organizations

share data

»» Taking advantage of data sharing

opportunities

Chapter  1
Getting Up to Speed on
Data Sharing Basics

E
very day, organizations everywhere use data to track busi-
ness results, make decisions, engage customers, define and
create products, forecast trends, and more. Data is also a
resource used and consumed between organizations, internal and
external to one another, to collaborate on business plans, mutual
initiatives, or joint opportunities.

There’s no limit to how enterprises can engage and collaborate

with data. However, data does not magically appear on your door-
step, figuratively speaking. It is generated at a place of origin and
then distributed across the organization and analyzed to gain
insights.

In this chapter, you learn about data sharing — what it is, why it
matters, how and why organizations share data, and what busi-
ness opportunities data sharing can create.

CHAPTER 1 Getting Up to Speed on Data Sharing Basics 3

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 What Is Data Sharing?
Data can originate from the many software applications an enter-
prise uses to run its business, from the constant activity of visi-
tors engaging a website, from an Internet of Things (IoT) device
attached to the refrigerator in your home, or from a sensor built
into something as sophisticated as the jet engine of an airliner.
There are potentially endless data-creating scenarios in the mod-
ern world. Market intelligence firm IDC estimates the world’s total
digital data created will increase to 180 zettabytes by 2025 (one
zettabyte is equal to about 1 trillion gigabytes). Unfortunately,
traditional data sharing methods require moving data, which is
riddled with problems. Going forward it will be impractical, if not
impossible, to share vast amounts of data in meaningful ways.

Furthermore, many enterprises have come to realize they could

enhance their business operations if they had access to data out-
side their organizations. Enterprises also recognize it is not easy
to access data they don’t generate themselves. Thus, data sharing
is the act of providing access to data between business units inside
the same organization, or between organizations external to each
other. The organization that shares its data is called a data provider.
The organization that wants to use shared data is called the data
consumer. Any organization can be a data provider, data consumer,
or both.

Figure  1-1 shows how organizations have traditionally shared

data — by making a copy of the shared data and sending it to their
data consumers. The data consumers then download the data to
analyze or combine that data with their existing data for deeper
insights into who their customers are, how efficiently their busi-
ness operates, and into which new industries their business is
heading.

But this process is slow, cumbersome, and costly and only allows
for moving limited amounts of shared data. Figure 1-2 shows how
modern data sharing happens without moving data. Instead, a
data provider makes available live, read-only copies of data to its
data consumers via modern cloud data sharing. In essence, data
doesn’t have to move.

Modern data sharing is a feature found in modern data platforms.

4 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 FIGURE 1-1: Traditional data sharing requires duplicating and moving data
from a data provider to data consumers.

FIGURE 1-2: Modern cloud data sharing enables fast, live, secure, and
governed data sharing without moving data.

CHAPTER 1 Getting Up to Speed on Data Sharing Basics 5

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 PORTAL FOR JOB SEEKERS
IMPROVES 300 PERCENT
Snagajob’s mobile sourcing and hiring tools connect 75 million
­registered hourly workers to Snagajob’s business subscribers, which
represent 300,000 employer locations.

Snagajob had a need to share its data from its data platform with an
external marketing analytics firm. The firm then used the data to reach
out to Snagajob’s business clients to execute targeted re-engagement
campaigns on behalf of Snagajob. This relationship allowed Snagajob
to avoid in-house labor costs associated with this marketing function.

To share its data, Snagajob routinely had to implement time-­

consuming steps that included:

• Identifying the database elements to be shared

• Extracting the data set with a client tool
• Compressing and encrypting the data in order to email it
• Emailing the file to the marketing partner
After the external marketing firm received the file, it would execute
the same time-consuming steps in reverse, with the additional steps
of building a database table to ingest the data and then importing the
data into its target database.

Snagajob turned to a modern, built-for-the-cloud data platform that

could operate as an extremely scalable data warehouse but also be its
platform for data sharing. All Snagajob had to do was create a “share”
that enabled its marketing partner to receive live, secure, and instanta-
neous access to the tables and views Snagajob shared. The marketing
partner could then execute the email campaign with data provided
that was always accurate and up to date, because the partner always
accesses a live, read-only version of Snagajob’s latest data. All of this
took place in a matter of minutes, without any data movement.

Performance, reliability, and agility were dramatically increased, allow-

ing Snagajob to reduce the implementation time for sharing data
from several days to just a few hours. Snagajob saved 300 percent on
costs and was able to operate much more quickly, which improved its
industry competitiveness.

6 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Data Sharing Examples
Here are just a few of the new business opportunities modern
cloud data sharing makes possible:

»» Data sharing to eliminate data silos: Develop a single

source of truth for all your internal data and share it among
thousands of data consumers across hundreds of business
units within a single organization.
»» Data sharing for business efficiencies: Share live data
with your business partners to optimize costs, streamline
operations, and provide superior customer service.
»» Data sharing as a product: Provide live and direct access
to slices of your data as a monetized service so your data
consumers can enrich their own existing data.
»» Data sharing as a product differentiator: Software-as-a-
service (SaaS) providers can offer direct access to the
petabytes of data generated from their business-to-business
(B2B) subscribers’ activity. Those subscribers can then
perform deeper analysis on more of their data — analysis
previously unavailable to them.

Capitalizing on these opportunities requires data sharing capa-

bilities with uncommon speed, power, security, governance, and
simplicity. These capabilities are not available with traditional
data sharing methods.

How Organizations Share Data

Traditional approaches require laborious efforts to stitch together
a patchwork of disparate tasks to share and move data. These
processes are costly, create manual overhead, and limit how much
data an organization can actually share:

»» Email: A data file is emailed from provider to consumer.

»» File Transfer Protocol (FTP): Data files are shared and
downloaded between two computers or via the Internet.
»» Extract, transfer, load (ETL) software: ETL software
extracts data from the provider’s database, transforms the
data, and then loads it into the consumer’s database.

CHAPTER 1 Getting Up to Speed on Data Sharing Basics 7

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 »» Online file sharing services: These are similar to FTP, but
sharing and downloading data files takes place via Internet file
transfer only.
»» Cloud storage: The provider stores data in the cloud and
provides the consumer with credentials for accessing it.
»» Application programming interfaces (APIs): An API is used
to initiate and manage the data transfer.

But imagine the possibilities of having on-demand access to

ready-to-use, live data so you can make immediate use of that
data inside a secure, governed environment.

Exploring Data Sharing Possibilities

With modern data sharing, the possibilities are practically end-
less. Here are a few compelling opportunities

»» Improve the customer experience: Delivering targeted

business or retail offerings with personalized marketing
campaigns in a highly competitive, digital market requires a
deeper understanding of your customers, competitors, and
industry trends. The primary path for gaining this under-
standing involves acquiring data you don’t already have in
order to reveal what you don’t already know.
»» Streamline your business: Easily sharing data across the
multitude of business units that comprise your organization,
and with your business partners, creates a single source of
truth. This could save billions of dollars by reconciling even
the most minor data inconsistences.
»» Create new business assets from data: Some data within a
data provider will be just as valuable to thousands of
external, non-competing data consumers. All of this can
happen through an effortless, self-serving business model
thanks to the simplicity of sharing any part of a modern
cloud data platform that offers modern data sharing.

8 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Looking at today’s data sharing methods

»» Recognizing the cost of delays

»» Looking at computing complexity

»» Addressing business pain points

Chapter  2
Understanding
Traditional Data
Sharing Challenges

I n this chapter, you learn about the many limitations of tradi-

tional data sharing methods and technologies.

Addressing a Multi-Faceted Problem

If you anticipate sharing data with tens, hundreds, or even thou-
sands of internal or external data consumers — each of which has
unique data sharing requirements — how can you easily support
this challenge? How do you support growth without constantly
building more storage clusters, managing complex software,
and  suffering through prolonged latencies and performance
­penalties — all without creating inconsistencies by sharing stale
copies of data? Simply put, the traditional data warehouse plat-
forms of today were not built to support the constant need to
share data in real time.

To understand the magnitude of the challenges associated with

traditional data sharing, consider the pros and cons of common

CHAPTER 2 Understanding Traditional Data Sharing Challenges 9

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 approaches (shown in Table 2-1) that a company would encounter,
for example, when sharing data with a third-party service pro-
vider or another external organization such as a business partner.

TABLE 2-1 Pros and Cons of Traditional Data

Sharing Approaches
Data Sharing
Approach Pros Cons

Email • Pervasive and • Not conducive for large data

ubiquitous sets from relational databases

• Infrastructure in (can’t scale)

place • Limited size of attached files
• Easy to compose (less than 25 MB) requires
large data sets to be decon-
an email and
structed and zipped
attach a file
• Limited network bandwidth,
which results in slow data
transmission

• Not secure, requiring custom

encryption

• Mirror effort required on

recipients’ end (receive,
decrypt, reconstruct data, and
so on)

File transfer • Well-known and • Schema changes require a

protocol long-established great deal of lead time
(FTP) — see protocol
• Must acquire FTP client
Figure 2-1
• Availability of a software, server, and/or service
wide range of FTP
client software • FTP account admin setup and
overhead required
and services
• Large data sets must be
deconstructed and broken
down in size to facilitate faster
data transfers

• Not natively secure; requires

custom encryption scripting or
secure service

• Mirror effort required on

recipients’ end (receive,
decrypt, reconstruct data)

• Efforts must be repeated with

each new update to a shared
data set

10 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Data Sharing
Approach Pros Cons

Extract, • Large availability of • Latency emerges when data

transfer, load well-established changes
(ETL) software ETL software
solutions
• Expensive software, costing up
to tens of thousands of dollars
• Purpose-built to • Complex, requiring specialized
extract data from skills to integrate and deploy
a database or data
source and • Can take months to
transform the data implement
for loading into a
target database
• Change management and
schema evolution can be
• Good for bulk and difficult
complex data
movement and
transformations

Online file • High availability • Better suited for sharing of flat

sharing of services files, not relational database
services
• Generally easy objects
to use • Data is not ready to use (ready
to analyze)

• Risk associated with data

inconsistencies when the
original copy of the data
changes

Cloud storage • Numerous • Less-than-optimal performance

services available when querying directly from
from large cloud cloud storage
storage providers
• Change management and
schema evolution difficult,
requiring a separate metadata
management process

• Risk exposure when data

changes

• Complete SQL data manipula-

tion language (DML) semantics
(for example, UPDATE, INSERT)
may not be supported

(continued)

CHAPTER 2 Understanding Traditional Data Sharing Challenges 11

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 TABLE 2-1 (continued)

Data Sharing
Approach Pros Cons

Application • Numerous APIs • Data movement is required,

programming available creating risks for failed
interfaces
• Wide variety of transfers
(APIs) use cases • APIs process data in small
• Programmatic amounts, creating bottlenecks
for large data volumes
implementation
relieves some • Performance is directly
manual effort affected by available band-
width, requiring high costs for
higher bandwidths

FIGURE 2-1: Multiple steps of a typical legacy FTP-based data sharing workflow.

Dealing with Conventional Data Sharing:

Time-to-Value Delays
Conventional data sharing methods can create other challenges
that cause more delays and require more assistance from your
IT teams, including:

»» Handling increased data size: The shared data set is often

much larger than originally scoped, which creates problems
with the data extraction process. You’ll likely need a scripting
language to automate the breakdown and extraction
process, which may require additional IT assistance. The
reverse process must also occur for data consumers.
»» Decrypting sensitive data: If the data set includes sensitive
information, the output files will likely need to be encrypted,
masked, or redacted, which may require additional IT assistance.
If the data set is encrypted, encryption keys must be securely

12 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 shared between the parties via a separate process, and the data
consumer must decrypt the shared data.
»» Changing file formats and schema: It may be necessary to
change the file format multiple times if additional database
attributes must be shared. When table attributes change on
the data provider’s end, a corresponding change must also
occur on the data consumer’s end.

The accumulation of all these steps results in slow and painful

processes for both data providers and consumers. All of this must
happen before any attempt to analyze and develop insights from
the data, which creates time-to-value delays.

Usually, the delays and difficulty don’t end with just the data
transfer effort. For example:

»» Sharing data in real time: More IT assistance is needed if

the data set is shared in a more real-time fashion, rather
than being sent only once per night.
»» Cleaning data: The import process has problems, and the
data isn’t as clean as anticipated. For example, the data
extraction may contain special characters that should have
been disregarded. This means the data provider must build
more-sophisticated data extraction processes, resulting in
more IT assistance, costs, and delays.

SPEEDING ACCESS TO DATA FROM

BILLIONS OF DEVICES
Localytics is a Boston-based company that provides a mobile
­engagement platform used in more than 37,000 apps on more
than 2.7 billion mobile devices worldwide. Localytics gives hundreds
of the world’s top brands insights about their mobile users and the
tools to engage with those users.

Localytics uses modern data sharing to provide its customers with

access to Localytics’ data without exporting that data, solving one of the
biggest data challenges marketers face. Previously, users had to con-
nect different sources of customer data from customer relationship
management (CRM) systems, business intelligence (BI) tools, mobile

(continued)

CHAPTER 2 Understanding Traditional Data Sharing Challenges 13

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 (continued)

analytics, and other sources. Data was often exported and c­ opied into
other platforms, manipulated, and further analyzed. The process pro-
duced multiple copies of the data living in many places, thus increasing
costs and complexity and producing inconsistent results.

ETL eliminated

Localytics removed the burden of cumbersome ETL efforts to make

data directly accessible through modern data sharing, creating a much
more efficient and reliable way to manage and understand customer
data. Specifically, Localytics employs secure, permissions-based access
to enable customers to work with session, event, and profile data from
Localytics and run their own queries and custom reports against that
data. Customers can also use popular BI tools to analyze their data.

Data latency reduced from three hours to three minutes

Localytics stores all its data in a modern cloud data platform,

­augmented with modern data sharing. Instant sharing of live data
with Localytics’ customers eliminated a previous data latency of three
hours. With modern data sharing, real-time data is ready to query
in about three minutes. Customers don’t need to expend effort to
use the data, which is filtered through live, secure, governed, and
permissions-based sharing.

Localytics customers immediately saw the potential to eliminate their

ETL processes, removing that burden from their overworked data
teams and saving their organizations time and effort.

To protect against failures during the file transfer process, on

either the extraction and/or import side, both the data provider
and data consumer must incorporate special software code or
scripts to monitor the transfer and automatically restart the pro-
cess in the event of failure. This means greater effort and longer
delays to develop insights and derive value from data.

Computing Complexity Challenges

Traditional options for sharing data also require scaling complex
computing platforms to share even small slices of data. Complex-
ity adds burdens and requires extra resources, including infra-
structure costs — internally and externally.

14 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 The goal should be effortless sharing of limitless amounts of data
with internal and external organizations, including your business
partners, for collaboration and business planning. If your busi-
ness model is focused on monetizing your data, you’ll want the
same level of effortless sharing to distribute data to as many data
consumers as possible, with individualized, self-service access
and security as needed.

If you think cloud storage is the answer, think again. Sharing data
using a basic cloud storage service is inefficient. It won’t provide
the ability for you or your data consumers to query the data in a
high-performance manner or ensure data consistency. A Hadoop
computing platform is not the answer either because of its inher-
ent complexities and complications.

Sharing Data the Conventional Way:

Business Pain Points
Cumbersome and complex data sharing methods combined with
costly and inflexible computing platforms produce headaches for
organizations that need to collaborate on data. In addition, the pro-
cessing overhead required to extract data from a ­traditional data
platform and transfer that data to other organizations delays the
value shared data provides. Additionally, every time data changes,
data extraction and transfer processes must be repeated because
shared data is always a static version and becomes stale immediately.

The inability to extract insight from data quickly is an inhibitor

to maximizing the commercial value from data. Data consumers
encounter delays in developing insights, which can lead to dissat-
isfaction with the data provider. Furthermore, because common
methods for sharing data can’t incorporate changes immediately,
data consumers risk executing analytics on incomplete data.
This can lead to less accurate analytics or faulty conclusions for
­business decisions.

Data sharing within an organization

Data sharing scenarios within an organization include:

»» Sales groups share data with finance groups to track sales

and revenue to forecast an organization’s performance.

CHAPTER 2 Understanding Traditional Data Sharing Challenges 15

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 »» Marketing teams monitor and analyze customer data to
predict behavior and align demand generation programs.
»» Different subsidiaries of an organization share data with
each other to better align their go-to-market plans and gain
more understanding of the separate areas of the business.

When functional groups within an organization cannot share data

effectively, data silos result, and business collaboration suffers.
Each group will maintain its own data warehouse or data mart — a
copy of some of the data from the corporate data warehouse. Data
silo and data mart sprawl ensue and create unnecessary burdens
for IT and data platform teams.

Data sharing business-to-

business (B2B)
Some examples of B2B data sharing scenarios include:

»» A hotel booking website shares reservation patterns and

trends with hotel properties to develop promotional and
pricing programs.
»» A grocery chain provides store sales data to suppliers to
ensure shelves are adequately stocked to meet demand.
»» Retailers share in-store sales data to fashion merchandising
so the hottest trends are always available.

Whether sharing data to external organizations or ­receiving shared

data from them, if organizations cannot collaborate on data they
are less efficient and run the risk of operating at a higher cost and
lower productivity.

Monetizing data
An example of monetizing shared data is a data service company
that gathers mobile phone location information and usage data
and then shares the information with advertising agencies and
marketing groups so they can execute highly targeted campaigns
to specific consumers.

Look for more details about data monetization in Chapter 3.

16 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Tracing the history of data sharing
in business

»» Looking more deeply at data sharing

scenarios and the value to the business

Chapter  3
Recognizing the Business
Value of Sharing Data

I
n this chapter, you learn how data sharing methods have
evolved in business, why data sharing is critical to any busi-
ness, how businesses share data internally and externally, and
how the cloud and software-as-a-service (SaaS) change the data
sharing model.

Looking Back at the Early

Days of Data Sharing
Understanding the business value of data sharing today requires
a historical perspective. Not long ago, it was considered the norm
for organizations to host and support multiple business applica-
tions within their own data centers. There would be an applica-
tion for finance, another for marketing, others for sales, human
resources, operations, and so on. Just ten years ago, large compa-
nies would host and run hundreds of business applications from
their own data centers.

CHAPTER 3 Recognizing the Business Value of Sharing Data 17

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Each of these applications would also have an associated data-
base. These databases were not optimized for analytics and did
not share data between applications. In order to analyze this data,
each business unit in charge of a database would have to extract,
transform, and load (ETL) the data into its own data mart, which
is a smaller, stand-alone version of a data warehouse. Then, to
develop business intelligence across an organization or to exe-
cute analytics against companywide data, data would have to be
sent through the ETL process from the individual data marts into
a central data warehouse. This data would then be prepared for
analytics. The entire process was slow and cumbersome. Data for-
mats varied across the applications, requiring further modeling
and transformation into a new data warehouse. But at least you
had access to the data because it was in your own data center.

The bottom line is that no company survives without some level

of internal data sharing.

Assessing the Business Value of Data

Sharing for Organizations
Data sharing across and beyond an organization consists of four
basic workflows:

»» Across lines of business (LOBs): Sharing data between

business units within the same organization
»» Between organizations: Outbound data sharing to another,
separate organization to benefit your business
»» Between organizations: Receiving inbound data shared
from another organization to benefit your business
»» Monetizing data: Sharing live data as a service so data
consumers can enrich their own, existing data

Sharing data across LOBs

Within the same organization, business units depend on email,
spreadsheets, shared network drives, application programming
interfaces (APIs), and other methods for communicating and for
sharing data. Sharing data across an organization enables and
fosters increased levels of business intelligence and drives timely
and informed business decisions.

18 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Within an organization, however, data is often locked in silos.
Mergers or acquisitions, firewall restrictions, or other business
or technology barriers often restrict an organization from easily
sharing data across its business units. These physical or logical
separations of infrastructure can prevent two or more business
units from accessing all available data within an organization to
deliver all-inclusive, data-driven insights. These data silos emerge
when an organization relies on a traditional, on-premises data
warehouse or a traditional data warehouse ported to the cloud.

Sharing between organizations:

Outbound
External data sharing takes place all the time. A vendor-­supplier
relationship, a partner relationship, a developer-producer
relationship — or any number of other business relationships —
all require two or more organizations to collaborate with data to
drive business. In Figure 3-1, the primary organization is sharing
data, outbound, to the partner organization.

FIGURE 3-1: An organization, acting as the data provider, shares data with

its supplier, the data consumer.

For example, in a vendor-supplier relationship with data sharing,

a supplier knows in advance when to replenish the stock of a par-
ticular item. Well-managed inventory also prevents overstocking,
minimizing the need to significantly reduce prices.

Sharing between organizations: Inbound

Increasingly, organizations engage outside service companies.
These contracted companies can specialize in logistics, shipping,
marketing services, or sales operations, just to name a few. For
example, a large retailer would collect massive amounts of dem-
ographic data about its target customers. The retailer would then
share this data as a data provider to a data analytics company.

CHAPTER 3 Recognizing the Business Value of Sharing Data 19

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 STREAMLINING INTERNAL
OPERATIONS WITH DATA SHARING
Rakuten Rewards is a global conglomerate that since 1997 has helped
shape the way people shop online, offering cash-back deals and shop-
ping rewards on the world’s largest selection of products and services.
Each company division or subsidiary has specific legal requirements
and permissions related to which data can be shared, creating a com-
plicated data sharing infrastructure within the company.

Not only was the internal process of sharing data cumbersome, but it
also prevented business units from accessing each other’s data sets
for making more-informed business decisions.

Rakuten chose a modern, cloud-built data platform that enables its

business units and subsidiaries to easily receive governed and shared
data within minutes and combine it with their own data sources for
even deeper insights.

As Rakuten Vice President of Analytics Mark Stange-Tregear

recounted, the resulting transformation of internal data sharing
“not only allowed us to easily share data with groups outside our
legal umbrella but also allowed other groups with specific expertise
or niche analytics skills to work on our data sets without us having
to hire outside resources, making us a more nimble organization.”

From there, the analytics company would analyze the data for the
retailer. It would then provide the analysis back to the retailer in
the form of an inbound data share, as shown in Figure 3-2.

FIGURE 3-2: The organization is the data consumer, accessing the data from
its outside data analytics vendor, which is the data provider.

20 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 In other scenarios, the organization contracts a service provider
to perform a function the organization chooses not to perform in-
house. In turn, the service provider generates data as a result of
that service — data that belongs to the organization, which is the
service provider’s customer. With inbound data sharing between
organizations, the data generated by the service provider is shared
with its customer. The customer then executes additional analyt-
ics to develop deeper insights and value from additional data gen-
erated outside its data center but within its business ecosystem.

Monetizing data
Data can also take on more significance today than just day-
to-day collaboration. Data is a business asset  — a currency. As
such, data can offer different types of value depending on the
organization that wants to consume that data. Thus, as with any
asset, data has value. To monetize the value of its data, a provider
can sell data to consumers that can then use the data to advance
their own business objectives (see Figure 3-3).

FIGURE 3-3: An enterprise data provider creates new revenue opportunities

by sharing data with other organizations acting as data consumers.

CHAPTER 3 Recognizing the Business Value of Sharing Data 21

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Data consumers can use shared data without having to capture
and collect it themselves. They can benefit directly from analyz-
ing that data or combining it with other data to enhance its value.

But harnessing the value of data — either for consumption, mass

collaboration, or value-added business opportunities — requires
an easy method for enabling data access without actually moving
the data. Traditional data sharing methods are too costly, risky,
and labor intensive. Monetizing data requires the ability to eas-
ily and affordably create a secure, self-service business model to
share data between data providers and data consumers.

SHARING DATA TO HELP SERVE

90 MILLION GAMERS
PlayFab provides back-end services and data logistics for its game
studio customers that serve 75 million monthly average gamers and
15 million daily average gamers. PlayFab shares player data for each
game with each respective game studio. PlayFab also anonymizes
aggregated data across all studios and shares those results with all its
game studio customers. On the receiving end of the shared data,
game studios optimize games based on the shared player data from
PlayFab.

When gathering data, and sharing it with game studios, PlayFab regu-
larly encountered challenges such as capturing the right data the first
time, getting data from client devices and moving it into the data pipe-
line, managing constantly changing data schemas driven by new
game data events, and reducing soaring costs due to moving and
transferring high volumes of game data. To solve these challenges,
PlayFab adopted a modern, cloud-built data platform.

With modern data sharing, PlayFab sets up secure, governed, and live
views of the data with each game studio under a straightforward,
self-service business model. This method avoids ETL entirely. Secure
and governed views guarantee each game studio’s data is truly iso-
lated and game designers get direct access to a direct feed of live
game data, without any custom import required.

22 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Exploring a modern data sharing
architecture

»» Allowing access to database tables and

views for data consumers

»» Looking at a modern data sharing example

»» Protecting sensitive data with secure

views

Chapter  4
Enabling Live Data
Sharing

I
n this chapter, you learn how a modern, built-for-the-cloud
data platform architecture helps data providers and data con-
sumers overcome traditional data sharing challenges. You also
learn how to use real-time data sharing from inside a modern
cloud data platform environment to quickly and easily enable
secure and governed views of live data for your data consumers.

Exploring Modern Data Sharing

As discussed in Chapter 2, data sharing is a multi-faceted chal-
lenge. But traditional methods of data sharing fundamentally
address only one part of the challenge  — providing data con-
sumers with access to a provider’s data. Although traditional
data warehouses and data lakes (repositories that store massive
amounts of raw data until needed for analysis) were designed to
make data usable, their underlying architectures are not capable
of modern data sharing — that is, providing data access to data
consumers without having to move the provider’s data.

Traditional data sharing is slow, and it reduces an organiza-

tion’s ability to execute quickly. In addition, a lack of security

CHAPTER 4 Enabling Live Data Sharing 23

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 and ­ governance, among other things, means traditional data
­warehouses and data lake architectures cannot support unlimited
concurrent access by data consumers or real-time data changes
by data providers without cumbersome unloading and transfer-
ring of data, as shown in Figure 4-1. This puts data consumers at
risk of operating on stale (static) data.

FIGURE 4-1: Traditional data sharing requires cumbersome, multi-step

processes from both the data provider and its consumers.

The lack of a comprehensive solution creates a struggle for data

providers and consumers to easily share data and ensure data
consistency. It also limits the ability to monetize data.

Winning with modern data sharing

With modern data sharing inside a modern, cloud-built data
­platform, in a matter of minutes you (as a data provider) can ena-
ble live access to any of your data for any number of data consum-
ers, inside or outside your organization. You can share data across
internal business units, with business partners across your eco-
system, and with external organizations to easily support richer
analytics, data-driven initiatives, new business models, and new
revenue streams (see Figure 4-2).

FIGURE 4-2: Providers can improve business with modern cloud data sharing.

24 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 With modern data sharing, ready-to-use data is ­ immediately
available in real time. In a modern cloud data platform architec-
ture, query speeds on shared data are exponentially faster and
fortified with limitless storage and compute resources. Modern
data sharing extends the architecture and functionality of the
modern cloud data platform to share data. Enterprises can grant
read-only access to their live, ready-to-use data (structured and
semi-structured) in a secure and governed environment. Data
consumers can then choose to combine (JOIN) data from other
organizations to augment and deepen their data analytics.

Only the scalability, elasticity, and flexibility of a multi-tenant

cloud data platform makes it possible to store data from diverse
sources and share that data among a large number of data con-
sumers without contention or competition for resources.

Making data sharing easy

Enterprises can realize the business benefits of modern cloud data
sharing thanks to a cloud-built data platform, which

»» Eliminates movement and copying of data: Modern data

sharing offers direct, real-time access to live data in a secure,
managed, and controlled environment.
»» Provides ready-to-use data: Data consumers get the full
capabilities of a data warehouse, allowing them to query and
analyze shared data as soon as they’re given access to it.
They can combine shared data with their own data. Security,
governance, data schema, and metadata are all provided
within the modern data platform.
»» Protects personally identifiable information (PII) and
complies with industry requirements: Data providers
can easily and securely share data while ensuring no PII or
sensitive data is compromised by using advanced sharing
functions, while still enabling the data consumer to make full
use of the data. In the same way, data consumers who are
interested in sharing data for enrichment by a data provider
can do so without exposing PII.
»» Enables data sharing without added costs: Modern
data sharing eliminates the duplicative costs of building the
infrastructure needed to store shared data, since data
consumers view the shared data directly from the data
provider without having to copy or move data.

CHAPTER 4 Enabling Live Data Sharing 25

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 »» Enables data sharing with unlimited data providers and
consumers: A modern cloud data platform can serve an
unlimited number of data providers and consumers, with full
transactional integrity and data consistency.

Modern cloud data sharing eliminates the delays, cost, and fric-
tion of existing methods, which provide only primitive mecha-
nisms for data publishing, access, and control. Modern data
sharing is built on three key architectural innovations, which we
discuss in the next few sections.

Decoupling of storage, compute,

and services
The separation of storage and compute resources is a funda-
mental part of a modern data sharing architecture, as shown in
­Figure 4-3. All data is stored, in optimized form and without any
loss of data fidelity, in the cloud. A single copy of the data stored
in a modern cloud data platform — a single source of truth — can
be accessed concurrently by any number of independent com-
pute clusters, enabling an organization to perform any number of
internal workloads, such as analytics.

FIGURE 4-3: A modern data sharing architecture built for the cloud with
storage, compute, and services completely separate but logically integrated.

Decoupling of storage and compute is also critical for sharing

data. It enables data consumers to directly access shared data,
using their own data platform compute power. But data consum-
ers don’t pay for storage costs (because the shared data doesn’t

26 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 move), and the data provider doesn’t pay for any of the compute
resources that a data consumer uses to analyze shared data.

Managing multi-tenant transactions

Making shared data usable requires access to data and coordi-
nation across all data consumers to ensure consistency, security,
and performance. The services layer is a key part of a modern data
sharing architecture. Global metadata, transactions, and security
are all managed from here, making the services layer the control
tower that tracks, logs, and directs access to data for every data-
base element and object contained within the data platform, as
shown in Figure 4-4.

FIGURE 4-4: Metadata in a modern data sharing architecture enables live data

access between a data provider and data consumer, without moving data.

Additionally, the services layer provides transactional consistency

across all data providers and consumers, ensuring all data users
see a consistent view of live and up-to-data data. A data provider
can update shared data in real time. Likewise, after transactions
are committed, all data consumers can simultaneously view the
data provider’s updates and immediately query the shared data —
all with transactional, ACID-based consistency.

ACID is a consistency model that defines a set of properties to

ensure transactions in a relational database are valid, even in the
event of multi-statement transactions and processing errors, as
well as power failures and crashes. The properties of ACID are:

»» Atomicity (“all or nothing”): Every operation in a transac-

tion must succeed for the transaction to be completed.

CHAPTER 4 Enabling Live Data Sharing 27

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 If even a single operation fails, the entire transaction is rolled
back and the database state is left unchanged.
»» Consistency: The completion of any transaction brings the
database from one valid state to another valid state.
»» Isolation: Concurrent transactions do not contend for
access to the data and are run as if each transaction
executed sequentially.
»» Durability: After a transaction is committed, it remains
committed.

Data providers also benefit by having to share only the specific

data the consumer needs, which makes the process easier and
more secure.

Gaining unlimited concurrency

With modern data sharing, shared data can be accessed by large
numbers of concurrent data consumers, as shown in Figure 4-5.
In contrast, the architecture of traditional data warehouses forces
all users to compete for resources, creating a struggle to deliver
optimum performance and consistency. Automatic scaling of
concurrency takes simultaneous query processing even further
in modern data sharing by automating the scaling of additional
compute engines without manual intervention.

FIGURE 4-5: Unlimited concurrency with a modern data sharing architecture.

28 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Using Modern Cloud Data Sharing
Modern cloud data sharing allows access to database tables and
views for any user of a modern cloud data platform. When a data
provider shares data with a data consumer, the database object and
view are all from within the data provider’s environment.

In addition, a modern cloud data platform gives data providers

granular control of access to database tables and secure views via
shares. Data consumers can query a provider’s database only if
granted access privileges. Once the data provider creates a share,
the data consumer can then query the data.

Instant access, without data copying or movement, is made possi-

ble because all database objects are maintained and updated only
in a modern cloud data platform, and orchestrated by its global
metadata management services.

Learning How It All Works

As a data provider, the first step to sharing data is to specify what
database tables and views to share with specific data consumers.
This is done via a data share object, effectively an “empty shell”
that houses the references to the actual database and the shared
database objects. Data shares are first-class objects in a modern
cloud platform environment for which it provides a set of data
definition language (DDL) commands for creating and manag-
ing shares. Commands include create share, alter share, drop
share, and others. Access commands include grant and revoke
privileges. Once a share is created, the data provider grants access
to the specific database and database objects it shares. The SQL
semantics are as follows:

1. Create the share. The following example creates an empty

share named sales_s:
create share sales_s;

2. Add privileges for objects in the share. Grant usage on the

primary object before granting usage on any objects within
the primary object. For example, grant usage on a database
before granting usage on any schemas contained within
the database. Complete all grants for the data share before

CHAPTER 4 Enabling Live Data Sharing 29

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 adding the data sharing consumer(s). The following example
grants privileges for the sales_db database, the aggregates_
eula schema, and the aggregate_1 table to the data-share
object:
grant usage on database sales_db to
share sales_s;
grant usage on schema sales_
db.aggregates_eula to share sales_s;
grant select on table sales_
db.aggregates_eula.aggregate_1 to
share sales_s;

3. Confirm the contents of the share:

show grants to share sales_s;

4. Grant access to the share for the intended data consumer(s).

The following example makes the sales_s share available to
data consumers A and B:
alter share sales_s add
accounts=data_consumerA, data_
consumerB;

data_consumerA and data_consumerB now can see their

individual shared data and can create their databases from
the shared data as necessary.

Controlling Access to Shared Data by

Using Secure Views
What if you have sensitive data in your database? With modern
data sharing, you are not limited to sharing entire databases or
entire database tables. If portions of a table are subject to strict
security and confidentiality policies, sharing the entire table
exposes the sensitive data. With a command utility called secure
view, you can control access to shared data and avoid security
breaches, as shown in Figure 4-6.

For example, for online retailers to plan inventory levels, they

need to share merchandise and sales data with their distributors.
However, the table within the database that contains the sales
data also contains sensitive customer ID information, which must
be blocked and protected.

30 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 FIGURE 4-6: Secure views with modern data sharing allow data providers to
protect access to sensitive data.

To demonstrate how to accomplish this using a secure view, fol-

low the SQL semantics from the previous section to create a data
share object, sales_s for database sales_db. For this example,
assume sales_db and sales_s already exist. The schema for
sales_db is named public, within which, table unitsales is
constructed as unitsales (customerid, sku, date, qty) and
is populated with data.

When you need to plan new inventory with your distributor, you
want to provide access to the unit sales data, but not the cus-
tomer  data, which is sensitive. Therefore, a secure view named
distributor_sales_data is created from the unitsales table,
just for the distributor.

With modern data sharing, the steps are accomplished as follows:

1. Create the secure view. Assuming the database and

schema are already created and populated with data, the
next step is to create a secure view on the unitsales table:
create secure view sales_db.
public.distributor_sales_data as
select sku, date, qty
from sales_db.public.unitsales;

CHAPTER 4 Enabling Live Data Sharing 31

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 This logic creates a secure view named, but without the
sensitive customerID data. The data included in the view are
sku, date, and qty.
2. In the sales_s share container, add privileges for the
secure view:
grant usage on database sales_db to
share sales_s;
grant usage on schema sales_db.public
to share sales_s;
grant select on view sales_db.public.distributor_
sales_data to
share sales_s;

This logic enables the share (container), sales_s to have

privileges for the distributor_sales_data secure view.
3. Confirm the contents of the share:
desc share sales_s;

The user is provided a readout confirmation of the share,

sale_s.
4. Grant the distributor access to the share (it is assumed the
distributor is also a modern cloud-built data platform user):
alter share sales_s add
accounts=<distributor_name>;

5. To see your share:

show shares;

As this demonstrates, data providers can easily share data while

controlling data consumers’ access to data with a secure view.
Sensitive data is protected, and data consumers gain access to
non-sensitive data for their own analytics, without the need to
copy or move data.

32 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Collaborating on truly shared data

»» Leveraging low-cost compute power

»» Using analytics to drive faster business

decisions

»» Taking advantage of data

commercialization opportunities

Chapter  5
Assessing the Impact of
Modern Data Sharing

I
n this chapter, you learn how modern data sharing enables
real-time collaboration. You also learn which technologies and
trends have enabled a modern data sharing architecture, as well
as how modern data sharing can enable organizations to quickly
create new business assets from data.

Modern Approaches to Data

Collaboration
Traditionally, data sharing has meant sharing copies of data
(see  Figure  5-1). This creates a myriad of challenges, including
the following:

»» Multiple, static, and out-of-sync versions of the same data

exist in different environments, across multiple data silos.
»» No practical single source of truth or governance exists for
data in the organization.

CHAPTER 5 Assessing the Impact of Modern Data Sharing 33

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 »» Critical business decisions are made based on outdated,
incomplete, or inaccurate data.
»» Electronic discovery costs escalate when multiple sources of
data within and outside an organization must be identified,
searched, and produced for litigation support.
»» The potential number of data breaches and accidental data
loss/disclosure risks multiply, along with their associated
costs, such as breach notifications, credit monitoring
services, damage to an organization’s brand, customer
churn, litigation, forensic analysis, and recovery.

FIGURE 5-1: The difficulties of traditional data sharing methods.

Modern data sharing enables fast, cost-effective, and secure col-

laboration inside and outside the organization by providing data
consumers with real-time access to a single copy of the same
data. With modern data sharing, data providers can share data
using the same modern cloud data platform they use to run their
organizations.

The Democratization of Computing

Enterprise-grade computing power is everywhere, and it’s get-
ting even more powerful, faster, and less expensive every day.
The cloud has further accelerated this trend, making limitless

34 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 computing resources easily available at massive scale to any-
sized organization.

With this much power, business users perform many analyti-

cal computing tasks rather than submitting them as queued job
requests for IT administrators to perform. Business analysts have
the power to run advanced analytics against large data sets. Data
scientists can perform predictive analytics and develop machine
learning algorithms that serve as the basis for artificial intelli-
gence. Business executives can execute swiftly with up-to-the-
minute analytics dashboards; product management executives
achieve faster time to market for new products and services. In
short, massive computing power is now within everyone’s reach.

Data-Driven Business Decisions

To ensure the most complete view of customers, today’s organi-
zations require a multi-channel approach to gathering data from
various channels (such as websites, mobile devices, point-of-sale
terminals, call centers, and so on). Traditional data warehouses
not only fail when sharing data, but also perform poorly when
needing to uncover meaningful relationships between different
forms of structured and semi-structured data.

Semi-structured data, such as JSON data, does not conform

to the standards of traditionally structured data and includes
data ­generated from newer data sources, such as social media
sites, clickstreams, mobile devices, and Internet of Things (IoT)
devices.

But collecting and analyzing all this data from multiple channels
is exactly what successful organizations must do. Outcomes from
insightful analytics can help them better target new products
and services. So, it’s understandable those that want to provide
data as a service, or as a value-added business asset, are just as
interested in delivering access to data quickly and easily, so other,
non-competing organizations can benefit.

CHAPTER 5 Assessing the Impact of Modern Data Sharing 35

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 REDUCING COSTS AND
STREAMLINING OPERATIONS
Heap Analytics provides web and mobile analytics for companies
across a number of industries. To share raw data with customers, the
company would drop files into Amazon S3 buckets, or host files with a
cloud-based vendor for customers who weren’t sophisticated enough
to handle all the extract, transfer, load (ETL) activities required to then
analyze the data. Heap’s costs and efforts included:

Engineering work: Five to ten hours of work per week debugging

pipelines, performing cluster maintenance, handling unique customer
setups, and resolving resource contention

Customer service work: Five or more hours of solutions time per

week getting proper access; scheduling cluster operations; and creat-
ing, maintaining, and resizing hosted clusters with a traditional cloud
data warehouse

By leveraging a modern cloud-built data platform with modern,

secure data sharing capabilities, Heap eliminated the engineering
work and reduced customer service work significantly. It also saw
gains in other areas. For example, Heap can now share data to its cus-
tomers, who can access it immediately and see value sooner. The
modern solution decreased Heap’s support burden by requiring less
back and forth on permissions and security approvals, no custom set-
ups and less infrastructural variance, and painless cluster mainte-
nance, with no resource contention.

The Commercialization of Data

By removing traditional barriers of on-premises and cloud data
warehouses, modern cloud data sharing introduces at least four
new economic opportunities that enable enterprises to share data
as packaged and monetized assets, quickly and securely, power-
ing a true data economy:

»» Data monetization: Many companies produce and sell

data, some of which started nearly 100 years ago. Now, with
modern data sharing, any organization can turn its data,

36 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 regardless of its size, into a business asset by charging for
access to slices of its data repository. This low-cost, zero-
headache solution enables data companies to immediately
meet data consumers’ urgent demands for fresh data with
up-to-the-minute accuracy, which maintains the highest
value for data. Organizations have greater power and
platform capabilities to support, move forward, and
implement data monetization strategies.
»» Data sharing with business partners: Sharing data directly
with business partners is not new. But the effortless sharing
of live data is groundbreaking. Modern cloud data sharing
enables the following:

• Organizations can share data instantly with business

partners in their ecosystem — for example, supply chain,
distribution, marketing, third-party sales, and so on.

• Data consumers can query against live data to mutually

benefit both the data consumer and data provider, no
matter which organization owns the data and no matter
which one takes on the role of either the data provider or
data consumer.

• Business results can be delivered faster from faster

execution of analytics based on internal and external data
sources accessed immediately, without overhead, and at
a fraction of the cost of traditional data sharing methods.
»» Breaking down enterprise data silos: Modern data sharing
eliminates data silos strewn across an organization and beyond
by allowing you to store structured and semi-­structured data in
a modern cloud data platform. You can share seamlessly,
without downloading or replication. Your systems that were
previously siloed can be tightly integrated, without manual
integration or the need for data pipelines.
»» Near-zero management to reach more data consumers:
Only a modern cloud data platform eliminates the traditional
and time-consuming methods needed to manage a legacy
data warehouse. Performance is built into the modern data
platform, so there’s no infrastructure to tweak, no knobs to
turn, and no tuning required. With near-zero management
required, organizations can pursue more, far-reaching data
sharing strategies to target a larger base of data consumers
across their business units, with business partners inside
their ecosystem, and with other organizations as part of the
data economy.

CHAPTER 5 Assessing the Impact of Modern Data Sharing 37

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 THOUSANDS OF DATA
CONSUMERS — ONE
SOURCE OF TRUTH
PlaceIQ aggregates, collects, and anonymizes data from thousands of
applications on mobile devices. It then makes the data available for
companies that want to target and reach out to mobile consumers
based on their location and behavior. PlaceIQ customers may include
marketing companies, advertising agencies, and product producers,
to name a few.

One data source — thousands of self-service users: PlaceIQ cus-

tomers execute highly targeted campaigns based on the individual
usage and geographic movement of mobile device users. To do so,
each PlaceIQ customer may require a unique slice of the data to
­qualify potential prospects. The cost and effort of traditional methods
made it impossible to effortlessly deliver the data within a self-service
business model.

Effective business models — zero-data management: A modern

cloud data platform, built for data sharing, resolved the challenges
faced by PlaceIQ to scale and deliver thousands of individual data
consumer subscriptions with governed, secure slices of the data.
PlaceIQ now uses far fewer resources to manage its single source of
truth, while enabling its data consumers to self-serve their own data
subscription. With modern data sharing, PlaceIQ focuses on develop-
ing new use cases for its data that are differentiated and priced
accordingly for every customer.

Simplicity for PlaceIQ data consumers: PlaceIQ customers also

benefit by having a simple environment to assess shared data before
merging it with their own data sets. In doing so, PlaceIQ customers do
not have to sort through large amounts of irrelevant data to find the
information they desire. Plus, all the data from the modern data plat-
form meets common security and certification requirements. PlaceIQ
customers can blend data sets without compliance issues.

38 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Creating data exchanges to extend data
to third parties

»» Setting up real-time data sharing

»» Overcoming challenges with traditional

data marketplaces

Chapter  6
Using Exchanges
to Share Data and
Monetize Insights

E
xchanging data within a controlled ecosystem can yield rich
analytics and deep insights, leading to more-informed
­decision-making. In this chapter, you learn how your orga-
nization can participate in a data exchange to leverage third-party
data and create new data assets.

Examining the History of Data

Marketplaces and Exchanges
A data marketplace, also known as a data exchange, allows par-
ticipants to discover useful data sets. Traditional data exchanges
are accessible to participants through a portal or app store-like
environment. Many data providers offer a basic data set that is
free. They then charge for additional data sets and any data ser-
vices they may provide.

CHAPTER 6 Using Exchanges to Share Data and Monetize Insights 39

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Thousands of online marketplaces today link buyers and sellers.
Typically, the data vendor handles data transformation, prepara-
tion, and loading, while the marketplace is in charge of discovery,
collaboration, licensing, and auditing. It’s a huge opportunity: In
its September 2018 report, “Value of Data: The Dawn of the Data
Marketplace,” Accenture estimates the data-as-a-service (DaaS)
market is poised to reach $10.4 billion by 2021.

In contrast, a modern data exchange is built on the architecture

of modern data sharing, enabling one-to-one, one-to-many,
and now many-to-many data sharing relationships between data
providers and data consumers. It eliminates the cost and effort
associated with traditional data transfer, which plague most data
marketplaces.

Most organizations already share their data, or plan to do so,

via data sharing and their cloud-built data platform. There is an
immense and rapidly expanding marketplace for monetizing data.
In addition to earning additional revenue by sharing slices of your
data to data consumers, it’s an opportunity to forge new alliances
and glean new insights by combining your data with data and data
services from other providers. We look more closely at data mon-
etization in Chapter 7.

WHY USE A DATA EXCHANGE?

Use an exchange to provide and consume shared data and data
­services to:

• Create market opportunities: Tap new revenue streams by

charging data consumers for secure and governed access to read-
only versions of your data sets.
• Improve security and data access: Minimize the risk and cost
associated with copying data by opening secure, real-time
­connections to read-only data sets.
• Improve the customer experience: Give customers a fast, secure,
and cost-effective way to access data and data services, increasing
consumption and improving satisfaction for your data subscribers.

40 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Moving into the Modern Age with
Real-Time Data Sharing
Most companies derive value from a data exchange fairly quickly
as they leverage the deeper insights from additional data and data
services. Those data consumers acquire data from third-party pro-
viders in several ways, including via an API, using a self-service
interface for discovery and analysis, buying raw data via download
or file transfer, and using an app that reveals trends and insights.
After that, two common questions arise:

»» How do I scale my data-sharing efforts to include external

parties, such as customers and partners?
»» How do I find relevant data I can easily source from third
parties without bogging down my analytics operations?

The answer to both questions hinges on architecture. Modern data

exchanges don’t require data movement; extract, transform, load
(ETL) technology; or constant updates to share data with con-
sumers. There is no need to transfer data through File Transfer
Protocol (FTP) or to develop and maintain APIs. And since data is
shared rather than copied, no additional cloud storage is required.
Data providers can easily publish data and make it available for
instant discovery, query, and enrichment by data consumers, as
shown in Figure 6-1.

FIGURE 6-1: An architecture for an efficient, real-time data exchange.

CHAPTER 6 Using Exchanges to Share Data and Monetize Insights 41

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Establishing the Right Architecture
for Exchanging Data
Developing and maintaining APIs levies a heavy toll on data
­providers and data consumers. The providers must configure and
manage the APIs, while consumers have to establish custom links
to all marketplace vendors with which they want to do business.
Instead, the ideal data exchange provides the following mecha-
nisms for both providers and consumers:

»» Facilitate real-time access to live data sets without APIs or

ETL processes.
»» Upload structured and semi-structured data sets without
having to move or transform the data.
»» Create different views of the data and share these views securely
while protecting personally identifiable information (PII).
»» Provide data in a ready-to-use format to data consumers.
Some cloud vendors offer data exchange options to their customers
by leveraging the inherent capabilities of their multi-tenant SaaS
architectures. To ensure good performance in  these exchanges,
leading cloud vendors should offer a multi-cluster, shared data
architecture that separates compute and storage resources. This
allows data providers to incur almost no cost to share data and
allows data consumers to pay only for the compute resources, not
storage, to analyze the shared data.

To maximize convenience, an exchange should be free to join and

should allow participants to instantly publish data sets, ensuring
other authorized users of the ecosystem can always access up-
to-date data. The data storage layer should reside in a scalable,
pay-as-you-go cloud layer, such as Amazon Simple Storage
Service (S3), Microsoft Azure, or Google Cloud Platform (GCP).
The data platform provider wraps these storage assets in a ser-
vices layer to ensure everything is secure, properly maintained,
tuned, and optimized for self-service access. This type of archi-
tecture makes it possible to efficiently exchange data from one
centralized system, with dynamic elasticity. Data providers can
share an unlimited amount of data, yet recipients pay only for the
data and resources they actually use.

42 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 CREATING THE IDEAL DATA
MARKETPLACE
Weather data is an important and vital asset for business intelligence
across industries. In the U.S. alone, weather accounts for more than
$600 billion a year in lost revenue. Weather strategies help companies
forecast sales, mitigate risk, adjust transportation routes, and confi-
dently make decisions.

The Weather Source team has 75 years of meteorology and climatol-

ogy expertise. The company provides weather data from the year
2000 to present and forecasts of up to 15 days. Each of Weather
Source’s hyper-local solutions can be tailored to the points of interest
most relevant to a customer’s business.

Weather Source joined a data exchange to expand access to its data

delivery services. The data exchange, built on top of a secure data
sharing platform, offers an easy way for businesses to access data.

With the exchange, Weather Source is now able to reduce costs and
effort involved in publishing its data sets. Data transformation, load-
ing, and reconstruction are no longer required. New objects become
immediately available to all consumers, providing real-time data
access across the entire ecosystem. This more efficient operation
offers customers better access to Weather Source data in a SQL-
friendly format that’s always up to date.

If you want to ensure long-term flexibility and market expansion

for the data exchange you join, work with a cloud data platform
provider that can replicate data among multiple geographic regions
as well as store data in many different clouds — including AWS,
Microsoft Azure, and Google Cloud — all as part of one seamless
exchange. That way, you won’t be locked into one type of object
store or one particular cloud vendor as your data exchange evolves.

Some cloud data platform providers allow their customers to cre-

ate private data exchanges to share and monetize data and data ser-
vices within a self-contained ecosystem. Data platform customers
can leverage the exchange to set up new lines of business by mak-
ing their data and data services available to their end customers
and partners. Consumers can easily access these data exchanges,
paying only for data and data services they use, as per the terms
set by the provider.

CHAPTER 6 Using Exchanges to Share Data and Monetize Insights 43

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 GETTING HELP FROM YOUR DATA
EXCHANGE VENDOR
A modern data exchange should provide many unique benefits,
including:

• Access to thousands of data sets across industries and functions

• An easy way to search for and discover data for deeper analytics
• Access to live data sets without manual intervention
• Full access to granular and historical data
• No need to build and maintain ETL processes
• No added data storage costs because data isn’t moved or copied
• Access to value-add data services to enhance your data analytics

Older exchanges depend on FTP downloads, ETL transforma-

tions, and API connections, to name a few. Modern exchanges use
streaming data services or SQL queries to extract information,
without the need to transform or copy data.

44 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Discovering data monetization
opportunities

»» Evaluating governance and distribution

models

»» Pricing your data as a service

Chapter  7
Monetizing Your Data

D
ata monetization is a process by which a data provider
charges data consumers a fee to gain access to the provid-
er’s data or data services, so a data consumer can enrich its
existing data sets to benefit its business and its customers. Some
providers offer access to the data itself while others offer services,
such as data modeling, data enrichment, and data analytics.

Nielsen is a pioneer in this space, and its service is so pervasive it

has become a household name. For more than 90 years, the com-
pany has collected, analyzed, and sold consumer data to media
companies, advertising firms, retail organizations, and many
other industries. Nielsen’s proven business model has inspired
numerous niche data sharing opportunities. For example, in the
financial services industry, some companies collect stock mar-
ket data, package it, and sell it to brokers and hedge funds. In
the gaming industry, PlayFab is well-known for providing gam-
ing and analytics services for its live gaming studio customers
(consumers), powering more than 2,500 games that serve more
than 1 billion gamers. PlayFab provides its data consumers real-
time analytics, dashboards, and custom reports about their play-
ers’ activity, so studios can optimize their gamers’ experience and
drive more business.

CHAPTER 7 Monetizing Your Data 45

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Examining Industry Opportunities
Organizations can monetize data in countless ways, as shown in
Figure 7-1. For example, a telecommunications company can sell
location data to help retailers target consumers with ads. Con-
sumer packaged goods companies can share purchasing data
with online advertisers — or directly with customers. A logistics
company might sell data about transportation patterns and ship-
ping activity as an indication of economic trends. Transportation
agencies can collect data from tollbooths and bridges to optimize
navigation apps and minimize traffic congestion. From tracking
engine performance to monitoring human biometrics, data shar-
ing reveals previously unobtainable insights so organizations can
spot trends, make predictions, and take corrective action.

FIGURE 7-1: Data sharing unlocks valuable opportunities in every industry.

Acquiring data from other organizations likewise can enhance

internal operations. For example, customer service departments
may want external data about callers to segment prospects and
knowledgeably route opportunities. Marketing professionals look
to external data sources to target their messages and tailor their
campaigns. Sales teams collect and aggregate third-party data
about the organizations they sell to, so they can properly qualify
leads. Risk analysts gather data from cybersecurity experts to help
identify network intruders and circumvent fraud.

46 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Getting in Front of Data
Monetization Trends
Modern data sharing is growing every year as the volume of data
that organizations collect expands exponentially. A Forrester
Research survey of global data and analytics decision makers
revealed that more than 75 percent of respondents are expanding
their use of external data and nearly 50 percent of these compa-
nies already commercialize their data.

From Internet traffic to weather data, social media trends to

purchasing patterns, your data may be valuable to third par-
ties. Whether or not you are monetizing your data, chances are
that your competitors are. In IDC’s “2019 Predictions for Digital
Transformation,” the research firm predicted that 80 percent of
enterprises will create data management and monetization capa-
bilities by 2020, and that by 2023, 95 percent of entities will have
incorporated new digital KPI sets.

To determine where the opportunities lie, ask yourself these

questions:

»» Are your customers requesting access to specific data they

don’t have access to today?
»» Are they in search of more data, better data, better insights,
or new insights?
»» Are customers complaining about how you currently share
data? Is fulfilling their requests a burden to your staff?

Start by identifying the area of value for your customers. This will
help you determine what type of data and access to give them, and
how you might extend your current business practices to maxi-
mize future data sharing opportunities.

As you consider a data monetization strategy, consider these

options:

»» Selling access to raw data

»» Charging for specific data services (such as data enrichment,
data modeling, and so on)
»» Offering data analytics for a fee

CHAPTER 7 Monetizing Your Data 47

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Observing Good Data Governance
For all types of data sharing scenarios, you must observe all
government and industry regulations regarding the security
and privacy of consumer data, as shown in Figure 7-2. Regard-
less what data you plan to share, setting conditions and limita-
tions is important. Appoint a data governance champion to pay
attention to these legislative mandates; set internal policies and
procedures; respond to customer requests; and control access to
data using secure views, secure functions, and secure joins — as
described in the next chapter.

FIGURE 7-2: Make sure your data sharing practice complies with all pertinent
regulations and doesn’t compromise customer privacy.

Evaluating Distribution Channels

As discussed elsewhere in this book, sharing data requires a
secure, flexible business model that simplifies the exchange of
data between data providers and data consumers. There are
several potential distribution channels for selling data and data
services and building market momentum:

»» Direct to customers: The advantage of this approach is that

you’re selling to a well-known audience, with the chance to
deepen those relationships. The disadvantages are that it’s a
limited market, and you have to construct a dedicated
sharing platform.

48 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 »» Through a data broker: The advantage is ease of adoption
because the broker supplies a built-in clientele. The disad-
vantages include lack of control over the data and the
inability to forge direct relationships with users.
»» Via a data exchange. As explained in Chapter 6, a data
exchange is a scalable platform that supports flexible,
bidirectional data sharing. This is a great way to build your
brand and establish new types of direct relationships with
customers. The only possible disadvantage involves adop-
tion: If you’re starting with an early-stage ecosystem, it may
take time for the service to deliver on its full potential. In this
case, look for a data exchange that will help you generate
brand awareness and scale as part of its value proposition,
versus a traditional data marketplace where your data sits
there waiting for someone to find it.

Pricing Your Data Sharing Service

There are a number of ways you can bill customers and partners
for your data. For data that is streamed or continuously updated,
monthly subscription fees are common. However, if your data
set is relatively static, then most of the value is derived at the
initial time of consumption, with less value over time. In these
instances, it may be more cost-effective to ask customers to pay a
one-time fee rather than charge them for a subscription.

You can also consider offering tiered pricing according to differ-

ent levels of service:

»» Free: Select access to a limited data set

»» Standard: Full access to the data set
»» Premium: Full access to the data set plus mobile alerts,
basic insights, custom dashboards, and role-based reports
»» Elite: Premium plus predictive analytics and decision-
support tools

CHAPTER 7 Monetizing Your Data 49

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 DATA SHARING SERVICE OPENS
NEW REVENUE STREAMS
Maintaining product quality in the fast-moving consumer goods
(FMCG) industry requires rapid decisions related to inventory, ware-
house management, shipping, and delivery. By helping grocery suppli-
ers visualize trends in their data and share insights with retailers,
Atheon Analytics keeps goods moving to the right place at the right
time. Atheon’s SKUtrak service is the leading flow-of-goods tracker for
FMCG suppliers throughout the United Kingdom.

Fresh groceries are a byproduct of fresh data, and Atheon’s new data
sharing service, based on its cloud-built data platform allows the com-
pany to easily share data to SKUtrak customers, so they always see
the latest insights but without complex data-copy or data-moving pro-
cedures. “We provide data directly from retailers and share it with
suppliers and customers to make sure their products get to the
supermarkets where they need to be,” Atheon’s head of service deliv-
ery Rose Ahearne said. “The data platform helped with scalability and
created new product opportunities.”

Having a cloud-built analytics platform to securely share real-time

data is now an essential part of Atheon’s business. Through visualizing
data in SKUtrak, retailers and their suppliers can explore regular pat-
terns and anomalies to identify opportunities for improvement across
the entire grocery supply chain. This results in increased availability,
reduced waste, and better-met shopper needs. “Data sharing offers a
great new revenue stream,” Ahearne added. “It has become part of
our product portfolio. Developers and consultants aren’t waiting
around for data, and productivity has gone through the roof.”

50 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Complying with government and
industry regulations

»» Keeping sensitive data secure

»» Selectively sharing and controlling

access to a data set

Chapter  8
Governing Your Data

M
any organizations need to securely link their own data
with data generated by their partners, suppliers, cus-
tomers, and industry peers, but they have concerns about
protecting personally identifiable information (PII), protected
health information (PHI), competitive data, and other forms of
fine-grained data.

It’s a pressing issue: Today’s businesses must adhere to strict

regulations governing the security and privacy of consumer data,
such as the European Union’s General Data Protection Regula-
tion (GDPR), the United States’ Health Insurance Portability and
Accountability Act of 1996 (HIPAA), and the California Consumer
Privacy Act (CCPA). These regulations must be observed through-
out the entire lifecycle of your data — from creation and storage,
to usage and sharing, to archiving and deletion.

For the GDPR, as an example, there are three key areas to consider:

»» Principles related to how personal data is processed and

handled
»» The rights of individuals (data subjects)
»» Accountability, including being able to document and
demonstrate data compliance

CHAPTER 8 Governing Your Data 51

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 GDPR requires organizations to rein in the runaway ­complexity
that multiple data sources create and to establish a modern data
governance strategy on a global scale. Not only do most organiza-
tions keep their data in multiple locations (disparate data ware-
houses, data marts, servers, and clouds), but they also suffer from
hidden layers of data. This phrase describes the countless copies of
data that exist within organizations as a result of siloed teams and
departments sharing information.

Copies of data are rampant when organizations rely on older tech-

nologies, such as traditional on-premises data warehouses, which
require data to be exported and copied in order to be shared. And
if distributing data sets via cloud storage, File Transfer Proto-
col (FTP), or APIs is hard to track, what about employees who
manually share files and spreadsheets via thumb drives or email?
Without the right technology, fulfilling compliance requirements
is difficult, because when part or all of a data set is copied, you
have to apply your efforts to multiple data sets, possibly in mul-
tiple locations.

Fortunately, fulfilling data privacy and protection requirements

is significantly simplified when you have the right technologies
in place for sharing that data. Instead of dealing with any number
of data sets in any number of locations, you can establish a single
copy of your data, governed by advanced technologies for access-
ing and sharing it.

You can ease your compliance efforts by keeping a single “source

of truth” of historical and live data in a single location and grant-
ing on-demand access to governed slices of that copy of data.

To determine what type of data security you need, start by iden-

tifying the data sharing scenario that most closely matches your
use case. Throughout this chapter, we explain five such scenarios.

Sharing the Same Data with One or

Multiple Consumers
In this situation, the need is simple: You have a table or view you’d
like to share with one or many data consumers (see Figure 8-1).
For instance, if a retail chain wanted to share its entire data-
base with 50 franchises, it could use this method. The process is

52 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 relatively straightforward. First, the retailer would need to create
a share, a database object that grants permissions and data access
to the data consumers.

FIGURE 8-1: Using permissions to share the same data with multiple

consumers.

To complete the process, the retailer would add permissions for

all related database objects it wants to share, and then add the
50 franchise accounts to that share. This could be automated via
a scheduled script or through a workflow or provisioning system
using Python or SQL.

Sharing Different Subsets of Data

with Multiple Consumers
This scenario is more sophisticated because it involves sharing
different subsets of data with multiple consumers. For instance, a
car manufacturer might want to share data with its dealers about
production plans. Some dealers may be confined to a single state,
while others span many states. Using this method, the manufac-
turer could share the pertinent data with each dealer, regardless
of dealer location.

For example, Figure 8-2 shows data in one table broken up into

two groups. Group 1 contains row 1, which is visible only to Con-
sumer A. Group 2 contains rows 2 and 3, which are visible to all
consumers.

CHAPTER 8 Governing Your Data 53

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 FIGURE 8-2: How to share the different subsets of data with multiple
consumers.

Facilitating this kind of sharing requires a couple of additional

steps. First, the data provider needs to create an entitlements
table to match the dealer account names to the set of objects each
dealer needs to access. Second, the manufacturer must create a
secure view that controls the data each dealer can access when that
dealer runs queries.

Using Secure Views to Provide

Predefined Slices of Data
As explained in Chapter 4, if any portion of a table is subject to
security and confidentiality policies, sharing the entire table could
expose sensitive data. Secure views allow you to control access to
the data and avoid potential security breaches.

Secure views are an effective way to enforce table, column, row,

and even cell-level security when sharing data between organiza-
tions external to each other.

Data providers can each create secure views of their data and share
access to those views with other users of the SaaS cloud data plat-
form, even if these users are in other organizations. This is also
a useful method for organizations that want to grant database

54 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 access directly to multiple end customers. Secure views allow
these customers to see only their specific rows of data from each
table, but not the rows that contain data about other customers.

When should you consider using secure views? The following are
some guidelines:

»» To allow data consumers to access data without compromis-

ing security
»» When PII and PHI are involved
»» To restrict data access when multiple consumers need
access to the same database

Answering Questions While

Maintaining Data Privacy
Despite the power and flexibility of secure views, nothing prevents
users from running SELECT * queries and viewing or exporting all
the data made visible to them. To mitigate this type of access, you
can create secure shared user-defined functions (UDFs): small pieces
of SQL or JavaScript code that securely operate against raw data.
Secure shared UDFs allow the owner of the data to limit the spe-
cific types of questions and analyses that can be asked against
their data. It also lets a user ask specific questions of detailed data
without giving that user the ability to directly view or export the
raw data.

For example, imagine a retailer that wants to allow its suppli-

ers to see which items are commonly sold together with theirs, a
common practice known as market basket analysis. To enable this
type of analysis while preventing users from seeing the raw data,
a SQL statement can be wrapped in a secure UDF, with an input
parameter to specify the item number that is being selected for
market basket analysis.

Secure shared UDFs allow data owners to provide data consumers

with functions that can run across detailed data while preventing
the other parties from viewing or exporting the raw data.

CHAPTER 8 Governing Your Data 55

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Data providers can grant rights to multiple consumers to use their
secure shared UDFs, even if those users are in different accounts
or different organizations. Secure UDFs have become a great way
to guarantee data security and privacy, giving data providers and
consumers the following benefits:

»» Allows consumers to run queries without viewing the source

data or underlying logic
»» Limits the questions consumers can ask — in other words,
the types of queries they can run
»» Gives providers the ability to provide valuable data to
consumers without worrying about disclosing confidential
information or PII

Using Secure Joins to Combine

Data between Companies
When sharing data with parties outside your organization, a data
provider may find itself sharing data among two or more orga-
nizations that have common customers. This situation comes up
frequently between healthcare payers and providers, which each
maintain their own records about patients. How do you ensure
that when sharing data, each healthcare company can see data
about common patients only? How do you offer bidirectional
access to the data without revealing PII, as governed by HIPAA
guidelines? Can you exchange data about patients without actu-
ally sharing or transferring the names or any other PII or PHI of
these patients?

Secure joins allow you to jointly analyze only the records that are
in both of the data sets, and to prevent each data consumer from
seeing records about non-overlapping patients. For more details
regarding secure joins, see “Secure Joins: How to Join Data Between
Companies While Respecting PII,” at www.snowflake.com/blog/
secure-joins-how-to-join-data-between-companies-while-
respecting-pii.

A comprehensive data sharing service upholds security and com-

plies with government regulations by using secure views, secure
shared user defined functions (UDFs), and secure joins.

56 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Evaluating opportunities and defining
your organization’s role

»» Confirming your data sharing capabilities

»» Covering all the bases

»» Running a proof of concept and

executing for success

Chapter  9
Six Steps to Advance
Your Business with
Modern Data Sharing

N
ow that you understand the enormous potential of modern
data sharing and the challenges of traditional data sharing
methods, it’s time to consider the possible impact and
benefits of modern data sharing. This chapter outlines six key
steps to help you and your organization get started with modern
data sharing to advance your business:

1. Uncover data sharing barriers and opportunities. The

goal in this step is to gather a snapshot of data sharing
requirements within your organization — both now and for
the near-term future. You need a firm handle on data flows
and work processes already in place to share data.
When you have this information, focus on identifying the data
that has the potential to produce the most value. Ultimately,
the objective is to uncover laborious data sharing and data
transferring processes robbing you of productivity and
resources within your IT, data warehouse, and business
analytics groups. These barriers create delays in producing
business value from data. Identify the barriers and set the

CHAPTER 9 Six Steps to Advance Your Business with Modern Data Sharing 57

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 stage for easier and faster execution of your data sharing
business plans. Here’s what to look for:

• Data: What types of data must the data warehouse

contain? At what rate is new data created? How often will
data move into the warehouse? Will you, now or in the
future, want to also have easy access to a data exchange
via your data platform?

• Data flows: Identify the data sharing already taking place

within different groups of your organization and with
external entities. Map out these data flows to understand
who generates the data, who consumes the data, and
how that is being done.

• Work processes: Which teams currently manage the work

processes for data sharing or data transfers? Identify the
tools being used, and whether they are extract, transform,
and load (ETL) tools or data replication tools. This, com-
bined with the mapping of data flows, will help you
determine the current costs associated with data sharing.

• Future time-to-market and time-to-insight objectives:

Know your data sharing needs in the present and
near-term future, without regard to your current tools and
available expertise.
2. Define your role for each use case. Identify each current
and future data sharing scenario and whether your business
unit or organization will be a data provider, data consumer,
or both:

• Identify the organizations, internal and external, that must

be brought on board. All data sharing relationships have
at least two stakeholders — the data provider and the
data consumer(s).

• Identify and engage your data providers and consumers,

and confirm everyone has access to the capabilities of
modern data sharing across multiple geographic regions
and cloud infrastructure providers.
3. Confirm your data platform solution can easily and
cost-effectively enable modern data sharing. Look for the
following capabilities of modern data sharing:

• Data does not move: Modern data sharing enables data to

be shared without any data movement, ETL, or file
transfer. This is the lion’s share of the cumbersome work

58 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 required to share data. Eliminating data movement puts
you on a better path for limitless data sharing.

• Real-time updates: As a data provider updates its data,

every consumer should see the change as soon as it is
committed. This should happen without any extra work by
the data provider or consumer. Data consumers want this
level of integrity, which increases the value of your data.

• Individual secure views and secure functions: Modern data

sharing is not about simply creating access to your entire
data warehouse for all your data consumers. It’s about
having the granular control to easily provide the necessary
view of the data as required for each data consumer,
while protecting sensitive data and complying with strict
industry regulations, such as the General Data Protection
Regulation (GDPR) and California Consumer Privacy Act
(CCPA). Modern data sharing enables one-to-one,
one-to-many, and many-to-many data sharing relation-
ships. The tool you choose should give you this flexibility.

• Create Your Own Private Data Exchange: Easily create your

own private data exchange where you control who can
publish, share, and access the data.
4. Implement a proof of concept (PoC). After investigating
data sharing options, viewing demos, asking questions, and
meeting with vendor teams, you should execute a PoC
as soon as possible. A PoC is a process of testing a solution to
determine how well it serves your needs.
In addition, consider what else you can do above and beyond
what you do today with data. If you had a modern, cloud-built
data platform that enabled modern data sharing, what
additional business value could the system deliver? You may
want to monetize some of your data. How will you accom-
plish this? When setting up your PoC, list all current and
future requirements and success criteria.
For example, if your primary complaint about your current
data sharing methods is that queries take too long to run,
don’t focus solely on that issue. Your PoC should validate
assumptions about all or most high-value requirements,
including ease of migrating your data to the new solution,
loading new structured and semi-structured data, running
queries, and handling multiple workloads.

CHAPTER 9 Six Steps to Advance Your Business with Modern Data Sharing 59

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 5. Create an outreach plan. Outline the steps necessary to
engage your data consumers. Because modern data sharing
can facilitate one-to-one, one-to-many, and many-to-many
data sharing relationships, you must communicate the value
of your data and the benefits data consumers can expect to
receive.
6. Execute — demonstrate time-to-market and time-
to-value improvements to your business stakeholders.
After completing your PoC, demonstrate the benefits of
modern data sharing to your stakeholders. Estimate time
savings and related cost savings for your organization as the
data provider and/or the data consumer. Demonstrate the
improvements in productivity your organization will gain and,
if applicable, forecast the revenue potential for monetizing
your data. You should be able to develop a complete picture
of the ROI potential for modern data sharing and a modern
cloud-built data platform. You will then be well on your way
to taking data sharing to new levels of capabilities and
opportunities for your organization.

60 Data Sharing For Dummies, 2nd Snowflake Special Edition

These materials are © 2020 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
WILEY END USER LICENSE AGREEMENT
Go to www.wiley.com/go/eula to access Wiley’s ebook EULA.