Professional Documents
Culture Documents
DEFEND YOUR
RESEARCH28
Wall Street doesnt
favor firms that are
good at innovating
VISION
STATEMENT30
How vital are the
worlds great urban
centers?
COLUMN36
Ronald Coase on
saving economics
from the economists
MARKETING
Idea Watch
Know your
problem customers
Our study of five Fortune 1,000 firms revealed four distinct types of unprofitable
customers; the more they cross-buy, the
greater the loss. The proportion of each
group varied with the kind of firm.
Service demanders
Revenue reversers
Promotion maximizers
Spending limiters
Bad Apples
Who are these profit-destroying customers? Our study revealed four distinct profiles. Identifying customers who fit these
profiles is the first step toward neutralizing
their impact.
Service demanders. These people
habitually overuse customer service in all
channels, from phone to web to face-toface interactions. The more they cross-buy,
the more service demands they makeand
the more your costs rise. At the retail bank
we studied, requests from service demanders for things such as assistance with online banking and balance transfers more
than doubled after the customers began
cross-buying.
Revenue reversers. Customers in this
segment generate revenue but then take it
back. At firms selling products, this typi22Harvard Business ReviewDecember 2012
29%
27
B2B financial
services firm
67%
B2B IT
services firm
4%
58
hbr.org
3
22
24
31
24
Retail
bank
Catalog
retailer
12
47
51
Fashion
retailer
61
19
clude her from cross-selling campaigns. If companys toll-free number might be left
on hold longer times.
it determines that a customer is a spending
limiter, it might try to increase her spendUse the right metric. The most coming through upsellingfor example, up- mon metric for assessing cross-selling
grading a banking customer from a regular campaigns is the average number of differto a premium checking account.
ent products (or product categories) sold to
Demarket when necessary. Not all each customer. Wells Fargo publishes this
information in its annual reports and cites
customer cross-buying occurs in response
its desired cross-buy levelthe number of
to cross-selling; some customers cross-buy
on their own. When problem customers do, products it wants each of its customers to
firms might be wise to demarket them ownas a key strategic goal. Like the financial services firm discussed earlier, compato limit or terminate the relationship.
nies commonly use this metric to create
Ending bad customer relationships isnt
employee incentives.
uncommon; researchers have found that
But as weve demonstrated, cross-sellas many as 85% of executives in an array of
ing as an end in itself is a bad idea. Before
firms and industries have done so (see The
Right Way to Manage Unprofitable Custom- cross-selling to any customer or segment,
determine whether the effort is likely to be
ers, HBR April 2008). When firms have
profitable. Otherwise you may find losses
contracts with their customers, they can
sever the tie in writing. Sprint, for instance, mounting even as sales volume grows.
Remember our catalog retailer? When
sent letters to 1,000 of its problem customwe examined its customer data, we disers in 2007 canceling their wireless service
covered a group of problem customers
contracts because they had made what the
company deemed an unreasonable number who could have been identified two years
into their relationships with the company.
of customer-service calls. In the absence of
a contract, and particularly in a brick-and- The retailer had unsuspectingly nurtured
them, mailing them multiple catalogs and
mortar retail setting, ending relationships
engaging in other cross-selling efforts. And
can be trickier. In 2003 Filenes Basement
prohibited two sisters with histories of ex- regrettably, the promotions worked: Those
cessive returns and complaints from shop- customers bought 44% more items, on
average, over three yearsbut during that
ping at its stores. However, enforcing such
time the average loss per customer more
bans is obviously difficult, and they can
than doubled, resulting in an additional
draw negative publicity.
loss of $41 million.
A more feasible approach is to limit the
Before you launch your next cross-
resources devoted to problem customers.
For example, in 2003, after Comcast dis- selling campaign, pause and ask yourself,
covered that 1% of its broadband customers Do we really want these customers?
were using 28% of its bandwidth, it warned
HBR Reprint F1212A
those customers to limit their internet
Denish Shah is an assistant professor of
marketing, and V. Kumar holds the Lenny
use. (Many cable internet service providDistinguished Chair in marketing, at Georgia
ers impose caps on heavy users.) Similarly, State Universitys J. Mack Robinson College of
service demanders who repeatedly call a
Business.
December 2012Harvard Business Review23