The Digital Transformation Playbook: Rethink Your Business for the Digital Age

Step 4: Data Audit
Now that you have a strategy in mind, you need to assemble the data that 
it will require. That starts with surveying what data you already have that 
could be used to enable or power your strategy. You may have a large, estab-
lished data set based on your core product or service (like TWC). You may 
be starting with a data set on website visitors, or you may have access to 

---

116
T U R N D A T A I N T O A S S E T S
loyalty-program data. For some businesses, the only data may be an incom-
plete list of customer e-mail addresses.
Next you should identify what data you still need. For the purpose of 
the strategy you have sketched out, what data is still lacking? What will it 
take to provide the full view of the customer needed by your new initiative? 
You may need to increase your data in terms of
r
 more records or rows (e.g., expanding from a limited sample of your 
customers to a much broader list),
r
 more types of data (e.g., adding preference data and transaction data to 
your customer contact data), or
r
 more historical data (e.g., going back many months in time in order to 
develop an effective analytics tool that can model and predict future 
outcomes).
Lastly, now that you’ve identified the gaps, you need to determine ways 
to fill them. This is where you can apply the options discussed earlier: cus-
tomer value exchange, lead users, supply chain partners, public data sets, 
and purchase or exchange agreements.
Step 5: Execution Plan
For your data strategy to be effective, you must do more than assemble the 
right bits of data (the zeroes and ones). You must put that strategy to use in 
the work of your organization. The last step is to plan for the execution of 
the key pieces of your data plan.
What technical issues need to be worked out? This may include data 
warehousing, latency, or how quickly the data needs to be updated. Your IT 
people will need to weigh in here.
What business processes will need to change? Most data initiatives 
assume employees of your firm will make different decisions and take dif-
ferent actions based on your data. You will need to identify those changes 
in advance of rolling out any technical solution.
How can you test out your strategy and build internal support? One of 
the best ways is to integrate the new data strategy into an existing initia-
tive at your company. Jo Boswell, the program lead for Know Me at British 
Airways, knew that it would be difficult to enlist in-flight service staff if her 
initiative was seen as one more competing priority in their work. Instead, 

---

T U R N D A T A I N T O A S S E T S

117
she integrated Know Me with their existing customer service program, 
showing how its data would help staff to deliver on the same four “customer 
service hallmarks” that anchored all their training.
22
Data-driven strategies 
should be in line with everything your business is doing and help people to 
do their jobs better.
8
The Data Value Generator outlined in the previous five steps in an ideation 
tool; its goal is to enable you to generate multiple ideas for possible data ini-
tiatives in an area of your business. After developing these strategic ideas, 
you will need to test the assumptions behind each. Can you, in fact, get the 
data? Can you get buy-in from the business units in your organization to 
act on your findings? Will the results really matter to customers? Can you 
develop an initial pilot to test your data strategy for proof of concept? We 
will look in depth at the issue of how to iteratively develop new innovations 
like this in chapter 5.
Before we leave the discussion of data, though, let’s consider some of 
the challenges that a traditional, pre-digital-era enterprise may face in reor-
ganizing around data capabilities today.
Organizational Challenges of Data
When Mike Weaver was brought in as director of data strategy for the 
Coca-Cola Company, his mission was clear. “We must understand con-
sumers’ passions, preferences, and behaviors so we can market to them 
as individuals,” he told me. As an expert in the area of applied analytics, 
Weaver saw that this required building a data asset in an industry that is 
not traditionally rich in consumer data. By combining its MyCokeRewards 
loyalty program with a variety of other data sets—observed behaviors on 
its websites, social log-ins via Facebook, cookie stitching, and data from 
various partners—the company was able to advance rapidly toward its goal 
of becoming a more data-driven marketer.
But the biggest challenges, Weaver told me, were organizational, 
not technical. He compared the process of shifting business practices 
at “the world’s greatest brand/mass media company” to turning an air-
craft carrier at sea. He knew that the right data models could be used to 
develop advanced segmentation schemes for Coca-Cola’s customers, to 

---

118
T U R N D A T A I N T O A S S E T S
understand customers’ different needs and wants, and to allow the firm 
to better serve and communicate with them. But before installing all the 
data centers and analytics models that would allow for real-time targeting 
of customers, the company first had to plan out the changes to its busi-
ness processes. Before a brand can take advantage of its ability to differ-
entiate customer segments in real time and deliver targeted messaging to 
them, it first needs to learn how to create messages in a very different way. 
This kind of targeting doesn’t require Coke to create a single, blockbuster 
Super Bowl ad; rather, it has to create dozens of versions of the same 
message and test them to see which ones drive response among different 
customer segments. The first step of the journey, Weaver reiterated, is to 
plan the changes in your business process—before you start buying all the 
latest hardware or cloud services.
23
In my speaking, teaching, and work with a wide range of companies, 
I’ve observed a number of common organizational challenges that busi-
nesses face as they shift to a more data-driven strategy. Each of them is 
worth considering when developing a data strategy.