KNOWLEDGE IS POWER.

Data-driven marketing success is built on accountability.

2016
Part 1
A step-by-step guide to
data-driven marketing
TABLE OF CONTENTS

4 Dominate with data.

5 Change the way you market. Change the bottom line.

9 Take the right steps.

11 Walk before you run.

12 Know what not to do.

13 Remember that it’s all about the customer.

2016 | Knowledge is Power. 2

In a popular television series Half a century later in the real corporate world, the Done right, data-driven marketing is more than just
revolution heralded in that scene is finally hitting a tool you can use to boost sales. It’s a methodology
dramatising a changing the pivot point. Technology and creativity are fusing and a mindset that can dramatically increase your
to make marketing campaigns exponentially more company revenue and shareholder value. This three-
1960s America, a pocket effective than they were just a decade ago. part, step-by-step eBook series explains how to make
protector-equipped techie One real-life business guru, the late Peter Drucker,
it work for you:

wheels a massive punch-card said the goal of marketing is to know your customer • Part One (this eBook) covers the importance
so well your product sells itself. Today, marketers of accountability and attribution—knowing what’s
computer into an advertising can do just that, thanks to technology that makes working and what’s not—and what data-driven
it possible to collect, understand and use customer marketing can help you accomplish. It also explains
agency while a displaced data. With data-driven marketing—that is, mining how to get started.
creative department looks on, customer data and using it to inform marketing • Part Two discusses how to merge data from
decisions—you can now track a shopper’s entire disparate sources to create a complete view of every
none too pleased. journey. From the moment they discover your customer and use that understanding to deliver a
brand, through their purchases, and on to their loyal seamless customer experience.
evangelism, you can know in incredible detail if the
• Part Three explains how to use data and software
campaigns you’re aiming at them are working. And if
to predict what customers will do, and how to turn
they’re not, you can adjust them on the fly.
that capability into a business advantage.

2016 | Knowledge is Power. 3

Dominate with DATA.

There’s an old advertising saying often attributed to So far, there’s still a big gap between what’s possible The successful company also has to be able to
John Wanamaker, the 19th-century department store and what most marketing teams are actually doing. create a more fluid, compelling customer experience
pioneer. Supposedly, he said that half the money he The retail and financial services sectors are relatively than its competitors, and data makes this possible
spent on advertising was wasted—he just didn’t know advanced, according to Jeff Allen, senior director too. Your marketing teams must be able to look at
which half. of product marketing at Adobe.1 But in most other customer behaviour, instantly put it in context, and
industries, the level of sophistication varies widely immediately react with a customised response. They
Wanamaker died in 1922, but the problem of figuring
from company to company. That’s changing as have to be able to do this across multiple channels at
out whether marketing money was being spent
marketers recognise just how crucial technology once. Those channels include the ones we’re familiar
effectively persisted into the next millennium. Even
is to what they do. with—like websites, mobile apps and social media
during Madison Avenue’s heyday, the best a company
feeds—and others that haven’t yet gone mainstream:
could do was put a clever catchphrase or image out In a 2015 Adobe survey of more than 1,000 marketers,
smart cars, wearable devices and the increasing
into the world and hope for the best. If sales improved, only 20% described themselves as being ‘tech savvy’.
number of objects that connected to the web as part
marketers didn’t know for sure if it was the ad campaign But 52% said they believed the ideal marketer would
of the ‘Internet of Things’.
that made a difference or something entirely unrelated. fit that description—meaning a significant number
The result was advertising that often missed the mark. of those surveyed were aware they had a knowledge If a company can’t master these channels, its
gap to close.2 competitors will probably beat it to the punch. A
Then came the internet, and marketers could
2014 eMarketer survey of marketing leaders found
track clicks and page views. They could see what To truly compete today, you simply can’t remain in
that, over the next year, a full third planned to start
grabbed attention and start to understand customer the dark about campaign effectiveness. It’s essential
deploying technologies for website optimisation,
behaviour—but these radical changes were just the to know what works—and what doesn’t. If you
marketing automation, social media monitoring and
beginning of the digital marketing revolution. can’t attribute your successes and defeats to their
online campaign management. 3
true sources, you’ll have little to build on when
Today, many business leaders have yet to fully embrace
it comes to putting a data-driven marketing plan In the coming years, the degree to which marketers
the power of information. Their companies collect
into action. Fortunately, unlike Wanamaker and his embrace technology will make a significant difference
plenty of customer data, but use it only as a report card
contemporaries, we can now analyse customer in whether their companies succeed or fail.
to judge past performance. With the right investment
data from every point along the path that leads to a
in people and technology, they could do much more.
sale and beyond. That makes it possible to establish
Data is the key to happier customers, growing sales and
attribution, thus holding you accountable and
solving Mr Wanamaker’s conundrum. Using a data- Companies collect plenty of
demonstrating your contribution to business results.
driven approach, you can understand what really works. customer data, but use it only
as a report card ... they could do
much more.

2016 | Knowledge is Power. 4

CHANGE THE WAY YOU MARKET.
Change the bottom line.

So what can data-driven marketing do for your company? Five things: Until recently, marketers often relied on ‘last-click attribution’, a method ascribing a
sale to the customer’s last point of contact with the brand. For instance, if a customer
Understand the sale. received a promotional email and made a purchase within the hour, the email
campaign got all the credit. Intuitively, though, we know that the email probably didn’t
Data-driven marketing helps establish attribution—the process of giving appropriate
do all the convincing. The customer might also have seen a television ad, walked past
credit to each marketing action that contributed to a sale. When customers are
a billboard or noticed a company social media post. To really understand whether
targeted on multiple platforms, it can be hard to tell which communications ultimately
campaigns are effective, you have to take all those points of contact into account.
influenced the decision to purchase. Now, though, we can see and understand every
step of their journey. That may sound challenging, but it isn’t as difficult as it may seem. Today’s software
takes every interaction between company and customer into account, determining
In the early days of digital marketing, attribution models only accounted for a few
which was most influential. Businesses—at least the more advanced ones—have
data sources—such as the company website, display advertising and email. Today,
moved from last-click attribution to multi-touch attribution models that give credit
data sources may also include online audio or video, mobile web, mobile apps, social
where credit is due. Using fractional attribution, they give the right amount of credit to
media—even data from retail outlets and geolocation. As the number of platforms
each factor that influenced a customer action. And whereas most attribution models
available expands—say hello to your talking fridge!—attribution will become even
used to be based on simple rules, today’s more sophisticated models rely on statistical
more complex. Good analytical tools must be able to combine every source.
algorithms. When properly applied, these models show which campaigns are working
and weed out fruitless initiatives, giving management the opportunity to shut down
the latter instantly before spending spins out of control.
Today, data sources may also include online audio
or video, mobile web, mobile apps, social media—
even data from retail outlets and geolocation.

2016 | Knowledge is Power. 5

Prove your worth.
Data helps prove marketers’ contributions to the bottom line. Technology is changing Senior managers will get behind anything that lifts share price. They also tend to be
the role of the marketing department, so that increasingly it is seen less as a cost centre especially impressed when marketing data reveals untapped business opportunities.
and more as a moneymaker. With the latest attribution software, that role is evolving (Part Two and Part Three of this three-part series go into more detail about how
even further. to make that happen.) “Marketing is now actively engaged in business discussions at
the executive level because we bring data to the table,” says Rick Maddox, a senior
Consider the results of a Georgia State University study.4 V. Kumar and Denish Shah,
marketing manager for Adobe Analytics. “It’s foolish to argue with data.”5
marketing professors at the university’s J. Mack Robinson College of Business, found
companies can raise their stock price by using marketing analytics to determine a How do you bring data to the table? First, use it to show which efforts are most
customer’s expected lifetime value and then allocating marketing funds accordingly. effective and so deserving of recognition and continued funding. Second, as the
Georgia tech study showed, it can direct efforts towards discovered opportunities. And
They studied two Fortune 1000 companies: one business-to-business and one
third, it can hold marketers accountable to goals. One way to encourage accountability
business-to-consumer. Looking at five years of customer information, including such
is to ask teams to both look at their own performance and share it with others. If
data points as previous purchasing behaviour and responses to campaigns, Kumar
your team discovered that a well-timed email campaign led to an unusually high
and Shaw came up with a lifetime value for each customer—calculated as the net
number of conversions, that’s something to crow about, both to management and
present value of expected cash flows over three years. The companies then used that
colleagues who can build on what you learned. Establishing a shared document that
information to guide their marketing investments and the researchers observed the
can be collaboratively edited—in which teams share their achievements, failures and
company stock price over a nine-month period. The business-to-business company
insights—both educates and encourages everyone to meet performance goals.
saw a 33% jump in stock price compared with a 12% rise for its top three competitors.
The business-to-consumer company’s stock price rose 58%, compared with a 15% rise
for its top three competitors.

2016 | Knowledge is Power. 6

Make it automatic. For example, an accessories retailer might have a
segment made up of women under 40 who live in
Data-driven marketing allows for highly effective
London and recently downloaded the brand’s mobile
automation, which is pretty much what it sounds like:
app. Based on the behaviour of others in the segment,
the process by which software recognises a customer
the software might know that, after downloading the
and performs an action without human involvement.
app, the customer is likely to buy something online
“It could be sending them an email. It could be sending
within a week. The software could automatically
them a postcard,” says Adobe’s Jeff Allen. “It could
send out a prompt—say, an email announcing
be that, if you download the mobile app, a tailored
a discount—with a high chance of resulting in a
experience happens in the app.”6
handbag sale.
Successful automation is based on marketers’ ability Instead of basic segmentation,
There’s a clear correlation between marketing
to group customers into segments. Most marketers
automation and company revenue. 60% of companies the data lets you intelligently
are familiar with the concept of segmentation, or
with more than $500m in annual revenue deploy identify and target distinct and
dividing consumers into subsets that are likely to
marketing automation technology, whereas only
respond to particular messages or products. Data-
10% of companies making $20m to $500m use it.
actionable groups of valuable
driven marketing goes further, creating richly detailed customers.7
The percentage declines from there, with only 5% of
audience segments, helping determine the best
$5m to $20m companies using the technology, and
approach to each one and enabling rapid responses.
only 3% of smaller firms, according to the eMarketer JOHN BATES
It also lets the marketer know which customers a senior product manager
survey.8 The impact of data-driven marketing will
campaign should target, and just as importantly, Adobe
vary from company to company and campaign to
which to avoid.
campaign, but your software provider should be able
to create return-on-investment models to predict the
impact of new analytics tools.

2016 | Knowledge is Power. 7

Sell faster. Keep customers for a lifetime.
Data-driven marketing increases the speed at which your company can interact Lastly, customer data can improve relationship marketing—that is, marketing that
with—and sell to—customers. focuses on customer loyalty and long-term engagement. Making and maintaining
lifetime customers is fundamental for any forward-thinking business. The CFO of
Sometimes a response to consumer behaviour can occur almost instantaneously.
that accessories retailer should ask themselves not just how much the company will
For example, geolocation might tell a retailer that a customer who made a recent
make on the sale of a single handbag, but also how much it will make on sales of
purchase is three blocks away from its shop in London. Software instantly assimilates
handbags, shoes, wallets and belts to a single customer for years to come.
this information and sends them a push notification to visit the nearby location.
Data is the key to establishing this profitable, long-term customer bond. For starters,
Data can also increase the speed of ongoing customer communication to make sales
software can be used to spot shoppers who convert—turn from browser into
happen sooner. Based on data, the retailer knows detailed customer preferences. If
buyer—at an above-average rate. Once a group of these customers is identified,
a particular customer likes shoes but not handbags, automated marketing software
marketers can study their behavioural data like anthropologists in the field. Where
will highlight footwear when she visits the company website. If she never buys shoes
do they like to shop? How often? Which actions lead to a purchase? Combined with
that cost more than £100, the software won’t bother to show her more expensive
demographic data, answers to questions like these create a complete customer
pairs. The company doesn’t waste time—its own or the customer’s—crafting offers
profile. That profile can then be used to identify and target other groups with similar
she’ll never respond to. (How this rapid-response automation works is also covered
characteristics, automating whatever action is required—such as tweaking emails
further in Part Two and Part Three.) Meanwhile, the marketing team can use
or presenting special offers—and getting members of those groups to buy. Software
what it learns from the data to quickly craft and test an offer that will work.
also allows marketers to test specific strategies against one another to see which is
most effective.
We can really define our targets and segments
and filter out all the others that aren’t relevant.
We do customers a service and don’t damage
our company brand by being annoying.9

RICK MADDOX
principal product marketing manager
Adobe

2016 | Knowledge is Power. 8

Take the right steps. Step 1: set your sights.
In this moment of digital transformation, staying Throwing money at software without an agreed-upon goal is a recipe for frustration. For starters, the executive team
ahead of rivals means holding marketing campaigns should define its objectives. Most companies share at least a couple. They want to meet their measurable targets
accountable to their goals and being able to attribute on key indicators, such as conversion rates and account registrations. They want to provide a personalised, seamless
every customer action. If you can do that, you’ll have a customer experience across all channels. And they want to steer their business, rather than take a reactive approach.
solid foundation for a successful data-driven marketing
There are various ways to achieve these aims. Different businesses will make different choices about which channels
campaign, one that moves your company away from
matter, what information to collect, and how they’re going to measure and define success. The first, crucial step is
the gut-instinct calls of the past to decisions in which
getting everyone to agree on what they want to do.
you can have great confidence.

Getting started down this path may initially look Step 2: assemble your forces.
daunting, but it doesn’t have to be. The rest of this
Next, if a dedicated digital strategy team doesn’t exist already, your company needs to create one. Since effective
eBook—as well as the other two parts of this three-
digital marketing requires a variety of skills, team members might be drawn from all around the company. Once
part series—will show you how to do it. Breaking the
you’ve brought them all together, though, they should have clear-cut roles. The team should be responsible for
process down into bite-sized chunks makes it easier.
planning, creating and deploying content through digital channels; understanding and implementing analytics;
As Allen says, “If I’m missing lots of steps along the
developing and running tests; sharing insights; and working closely with colleagues in IT.
way, I can really get attribution wrong.”10

Step 3: take stock with an audit.

Perhaps most crucially, before getting started, your company needs to assess its current capabilities. This kind of
self-audit allows marketers to ferret out gaps in infrastructure and data gathering. In short, it shows how ‘mature’
the business is in terms of data use. Conducting regular audits is also a great way to check progress over time.

The best way to think about an audit is as a series of questions to ask about where your company stands now. You
could be starting from scratch or already well on your way, just figuring out how to make your efforts more effective.
Ultimately, an audit will serve as a blueprint for your move forward.

2016 | Knowledge is Power. 9

An audit in a nutshell. Decision-making Automation

Ready to launch a company audit? Here are some of How are digital marketing decisions made? At the What processes are handled automatically and which
the things you should be looking at as you grow your dawn of the internet, individuals or groups with access still require a slow-moving human hand? What’s the
data-driven marketing practice. to some basic data made marketing decisions based existing process for optimisation and how quickly can
on their opinions. As data-driven marketing practices the company act on data? At the most basic level, data
have become more sophisticated, more companies reports automatically go out to the staff members
Collection
are using technology to test hypotheses about what who need them. As a company matures, automated
What data is already being collected, where's it coming ads or placements are most effective. They make calls reports become more detailed and comprehensive,
from, and how is it being stored? As likely as not, your based on statistical analysis of customer segments providing, for instance, trend data and requested
company is already collecting some customer data— and automate some decisions in response to shopper real-time alerts. Ultimately, the goal is for automatic
maybe via its websites, existing customer relationship behaviour. They also test ideas, designs and initiatives systems to identify segments and rapidly optimise—
management software or its brick-and-mortar stores. against one another to figure out which ones work best. assimilating and responding to customer behaviour in
As your company matures, it will collect data from the most profitable way possible.
more sources (including third-party providers), merge
Attribution
and store it in one system, and use it to learn more
about its customers. Can your company link business results to specific
marketing efforts? If so, how? The simplest attribution
models look at just a single point of contact between
Analysis
the buyer and seller. More mature models test various
What are you already tracking and analysing? Volume experiences and use statistical algorithms for a better
of traffic? Ad impressions? Conversion rates? What understanding of which contact points influenced the
benchmarks are you applying and with what goals customer. By demonstrating what’s most effective,
in mind? As your analysis capabilities advance, you attribution models can also help you figure out how to
should be able to create very specific customer improve campaigns as your content evolves over time.
segments in order to understand and ultimately
predict customer behaviour.

2016 | Knowledge is Power. 10

WALK before you run. LET them see it.
Once you have a good sense of your capabilities, Once basic data is collected, crunched and put to Let’s face it. Most of us aren’t data wonks. For
the next step isn’t to seek out more data, but to good use, it’s time for the marketing department every staffer who loves staring down a column
make the most of what you’ve got. “Get the data to decide what additional information it wants to of numbers in a spreadsheet, you probably have
you have in order,” says Axel Schaefer, a senior collect. Two broad kinds of customer data need to 10 others who’d rather bench-press a filing cabinet
manager for strategic marketing with Adobe be brought together. The first—behavioural data— stuffed with spreadsheets. However sophisticated
Marketing Cloud.11 In other words, a company includes past purchases, responses to promotions, its underlying algorithms, if your data-crunching
should figure out what to do with what it already preferred channels and levels of social media software can’t provide great visualisation tools,
has before it does anything else. A small investment engagement. The second—demographic data— it will be of limited use. A compelling, intuitive
in training and technology can provide an almost includes age, income, education level, gender and dashboard—a digital display of key business
immediate leap forward. location. Combining the two can yield greater information—gives managers easy-to-understand
insight into what customers are willing to spend, snapshots of how their campaigns are doing,
The right technology allows marketers to instantly
which devices they use to make purchases and enabling them to make swift decisions without
see how customers are behaving and then quickly
what circles of influence they belong to. needing a spreadsheet junkie around to translate.
adjust strategies and campaigns for maximum
impact. If a hotel chain can see that Tim Jones Brought together, all this data helps establish a A good dashboard shows the most important
visited its website and checked room rates in uniform profile for each customer that can identify metrics prominently and includes graphics on
Dublin, it knows a little. If it can see that he also him or her across every channel. If the hotel knows any variety of data that marketers might want to
downloaded the hotel’s mobile app while in Dublin, Mr Jones downloaded its app, but doesn’t recognise track. It should be customisable. One retailer, for
it has a clearer picture. If it knows that most people him as the same person who checked Dublin rates, example, may want to see how many total units
who check rates and then download the app while the pieces of the puzzle are much less valuable are being sold (and at what prices), which locations
travelling abroad go on to book a stay within three than if they were put together. (Part Two of this are performing better than others and how well
days, it has enough information to respond in a way series explains how to create a holistic view of every online sales are doing. Another might want to track
that could tip Mr Jones into making a purchase. customer, combining all the pieces of the puzzle how many purchases were made in response to a
for a complete picture.) With it, marketers can see television commercial for a specific sporting event.
It’s not just about how much the customer’s whole path to conversion—from “With the dashboard reports, you can figure out
data you have, it’s about keyword search to display ad to social media attribution,” Maddox says. “You can line up all your
promotion to advertiser’s website. campaigns, side by side, showing the investment
knowing what you have
and return on investment for each one.”13
and how to use it.12

AXEL SCHAEFER
senior manager
Adobe

2016 | Knowledge is Power. 11

KNOW what not to do.
Knowing what to avoid can be as useful as knowing Failure to collect mobile data is still a surprisingly Lastly, when companies fail to hire or assign enough
what to do. As you move forward, keep common common error. Optimising desktop analytics won’t skilled people to understand and manage data-driven
pitfalls in mind. help if most traffic comes from mobile devices. A marketing. It may seem obvious, but digital strategy
2014 comScore report found that 60% of all US digital doesn’t happen on its own. It requires dedicated full-
Keeping information segregated is a big mistake.
media time is spent on mobile devices. “Anyone time employees undistracted by other responsibilities.
“Often, audience data stays in silos across the
not measuring online activity in apps or in mobile The old staffing model—a lean in-house marketing
business,” says Pete Kluge, a product marketing
websites has the biggest issue,” Schaefer says.15 team working with agencies focused on art and
manager for Adobe Media Optimizer.14 Sometimes,
copy—is, quite simply, obsolete.
he says, different departments in the same company In the push to embrace data-driven marketing, some
have databases of customer information they don’t companies make the mistake of failing to get all the
share with one another. “It’s important to understand pertinent divisions of their business—marketing,
where all the audience data is across the business,” he sales, IT and management—on the same page
60%
of all US digital
notes. Databases—like those that track social media before they dive in. Every department has to be on
media time is spent
usage, loyalty and customer preference—must be board, and marketing and IT, in particular, will have on mobile devices.
consolidated. If that data is scattered across different to work closely together. “There isn’t always a holistic Source: 2014 comScore
departments and controlled by different teams, view across the business, so you have to bring the
its power simply can’t be harnessed. (Part Two marketing and IT people together,” Kluge says.16 (For
of this series delves further into why and how to more on getting teams to cooperate, see our eBook
merge disparate data sources.) IT and Marketing: Finding Harmony in the Yin and
Yang of the Digital World.)

2016 | Knowledge is Power. 12

REMEMBER that it’s all
about the customer.

It’s important to keep in mind that all the data, analysis and targeting is going towards one
goal: creating a smooth, enjoyable customer experience that grows your business. Consumers
now expect that every encounter with favourite brands will be personalised. Given the choice
between a company that doesn’t recognise them versus one that does so without a hitch
(no scrambling for passwords), many opt for the latter.

You can’t create that effortless customer experience without information. With the right
data and smart analysis, the marketing department gains the all-important power to hold
itself accountable and find out which efforts are responsible for which achievements. Once
you’ve established attribution, you have a springboard to make the big decisions on how to
engage and influence customers. Getting those decisions right will determine whether your
company stands out from the competition.

Part Two of this three-part series explains how to use data to create a
holistic customer view. To continue learning more about the concepts
covered here in Part One—like attribution, accountability, conducting
an audit and visualisation—check out this link.

To learn more about how you transform your marketing through data, visit
http://www.adobe.com/go/data-driven-marketing.

2016 | Knowledge is Power. 13

Adobe Marketing Cloud empowers companies to use big data to effectively reach and engage customers and
prospects with highly personalised marketing content across devices and digital touch points. Eight tightly
integrated solutions offer marketers a complete set of marketing technologies that focus on analytics, web
and app experience management, testing and targeting, advertising, audience management, video, social
engagement and campaign orchestration. The tie-in with Adobe Creative Cloud makes it easy to quickly
activate creative assets across all marketing channels. Thousands of brands worldwide—including two thirds
of Fortune 50 companies—rely on Adobe Marketing Cloud with over 30.4 trillion transactions a year.

1 Jeff Allen, personal interview, 9 June 2015.

2 Adobe, US Marketers’ Research Results 2015 (survey conducted by Edelman Berland), March 2015.
3 Tobi Elkin, ‘Marketing Automation: Examining Users and Usage, and What’s Coming Next,’ eMarketer, February 2015.
4 Francois Laxalt, ‘Can relationship marketing lift our stock price? Yes!’ Adobe Digital Marketing Blog, 3 November 2014, Accessed 14 September 2015.
5 Rick Maddox, personal interview, 25 June 2015.
6 Allen interview.
7 John Bates, personal interview, 26 June 2015.
8 Tobi Elkin, ‘Marketing Automation: Examining Users and Usage, and What’s Coming Next,’ eMarketer, February 2015.
9 Maddox interview.
10 Allen interview.
11 Axel Schaefer, personal interview, 23 June 2015.
12 Schaefer interview.
13 Maddox interview.
14 Pete Kluge. Personal interview, 24 June 2015.
15 Schaefer interview.
16 Kluge interview.

© 2016 Adobe Systems Incorporated. All rights reserved.

Adobe, the Adobe logo and Creative Cloud are either registered
trademarks or trademarks of Adobe Systems Incorporated in the
United States and/or other countries.

2016 | Knowledge is Power. 14