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

Targeting: Narrowing the Field
The second template for data value creation is targeting. By narrowing 
the field of possible audiences and identifying who is most relevant to a 

---

T U R N D A T A I N T O A S S E T S

109
business, customer data can help drive greater results from every inter-
action with customers. In the past, customers were often divided into a 
few broad segments for targeting based on factors like age, zip code, and 
product use. Today, advanced segmentation schemes can be based on much 
more diverse customer data and can produce dozens or even hundreds of 
micro-categories. How a customer is targeted can change in real time as 
well, as they are assigned to one segment or another based on behavioral 
data such as which e-mails they clicked on, rewards they redeemed, or con-
tent they shared. Ideally, customer lifetime value (as discussed in chapter 
2) should be included as one metric for targeting customers based on their 
long-term value to the business.
Custora is a data analytics company that helps e-commerce busi-
nesses determine the likely customer lifetime value (CLV) of their website 
visitors—that is, not just their likelihood to buy in this visit but their likely 
profit potential in the future. This is done by analyzing historical customer 
data and applying both a CLV model and Bayesian probabilistic models. 
For example, when a new customer makes just one purchase on a web-
site, Custora can predict that they are likely to make six purchases in the 
upcoming year, totaling $275 and placing them among the top 5 percent 
of the company’s customers. Other predictions based on historical data 
include the category the customer’s next purchase will likely come from 
(e.g., home furnishings vs. lawn care). The model can even provide warn-
ing signs—such as predicting that if this customer doesn’t place an order 
for three consecutive months, the business can assume they have only a 10 
percent chance of returning.
15
InterContinental Hotels Group carefully uses data on the 71 million 
members of its Priority Club loyalty program to understand and target 
them more effectively. This data includes much more than zip code and 
hotel room preferences. Up to 4,000 different data attributes—such as 
their income level, their preferred booking channel, their use of rewards 
points, and whether they tend to stay over weekends—are used to assign 
each member to a customer group. This level of segmentation has allowed 
the hotel to shift from sending out a dozen varieties of an e-mail mar-
keting message to sending out 1,552 different variations, targeted around 
past behaviors and special offers such as local events. These new market-
ing campaigns have generated a conversion rate (the portion of customers 
accepting the offer sent) that is 35 percent higher than that of less targeted 
campaigns the year before.
16

---

110
T U R N D A T A I N T O A S S E T S
Using data for targeting can even have a powerful impact in a field 
like nonprofit health care, thanks to a practice known as “hot spotting.” 
Dr. Jeffrey Brenner, a family physician in Camden, New Jersey, studied 
medical billing records from hospitals in his hometown and discovered 
that 1 percent of the town’s population was responsible for 30 percent 
of its health-care costs. “A small sliver of patients are responsible for 
much of the costs, but we really ignore them,” said Brenner.
17 
He used 
that data, and small grants from philanthropies, to start the Camden 
Coalition of Healthcare Providers and focus on “spotting” these patients 
and improving their care. Over three years, the organization was able to 
reduce emergency room visits by 40 percent among the initial group of 
the “worst of the worst” patients and to reduce that group’s hospital bills 
by 56 percent.
18