read about 30 page readings on business(overconfidence) then write a 500-600 words answering some questions excluding intro, conclusion, or background.
c46260823
writeup1read3.pdf
Want to Pertect Your Company ' �
service? Use Behaviora I
Science
The next frontier in service management comes
from the venerable field of behavioral science,
where provocative psychological research sheds
light on how customers feel when a company
"touches" them. The take-away.· five new
operating principles.
by Richard B. Chase
and Sriram Dasu
)UN[ 2001
WHAT DON'r WE KNOW about service management? For the past 15 years, legions of scholars and prac
titioners have studied the subject. They've applied queu-
ing theory to bank lines. They've deified well-run call
centers. They've measured response times to the tenth
decimal point They've built cults around "moments of
truth;' "service recovery;• and "delighting the customer:·
It may appear, then, that no stone in the service
management garden has been left untumed, not to men tion analyzed, polished, and replaced. Surprisingly little
time, however, has been spent examining service encoun
ters from the customer's point of view. Specifically, prac
titioners haven't carefully considered the underlying
79
Want to Perfect Your Company's Service7
psychology of service encounters-the feel
ings that customers experience during
these encounters, feelings so subtle they
probably couldn't be put into words.
end on an Uptick
Fortunately, behavioral science offers
new insights into better service manage
ment. For decades, behavioral and cogni
tive scientists have studied how people
experience social interactions, form judg
ments, and store memories-as well as
what biases they bring to bear on daily life.
Their findings hold important lessons for
the executives who design and manage ser vice encounters. First, the research tells us
a lot about how customers experience the passage of time: when time seems to drag,
when it speeds by, and when in a sequence
of events an uncomfortable experience will
be least noticeable. Second, it helps us un-
Dan1el Kahneman, a professor of psychology at Princeton University, is a
leadmg researcher m cogn1t1ve psychology. In a 1993 experiment, he and his
colleagues asked subjects to choose between two unpleasant experiences.
In the first, subjects immersed their hands in uncomfortably cool water (St> F) for 6o seconds. In the second, the same subjects Immersed their hand in cool
water (57° F) for 60 seconds followed by 30 seconds in slightly warmer water (59° F). Even though the second sequence extended the total discomfort time,
when subjects were asked which experience they would repeat, nearly 70%
chose the second one.
Kahneman found similar results in a field experiment he performed with
D.A. Redel meier. They learned that prolonging a colonoscopy by leaving the
colonoscope in place for about a minute after the procedure was completed
thus decreasing the level of discomfort for the final moments of the proce
dure- produced significant improvements in patients' perceptions of the
expenence.
derstand how customers interpret an event
after it's over. For example, people seem to be hardwired
to blame an individual rather than a poorly designed sys
tem when something goes wrong.
In this article, we'll translate findings from behavioral
science research into operating principles for service
encounter management. And we'll show how managers
can optimize those extraordinarily important moments
when the company touches its customers-for better and
for worse.
Applied Behavioral Science
In any service encounter-from a simple pizza pickup to
a complex, long-term consulting engagement-percep
tion is reality. That is, what really matters is how the cus
tomer interprets the encounter. Behavioral science can
shed light on the complex processes involved in the for
mation of those perceptions. In particular, it can help
managers understand how people react to the sequence
and duration of events, and how they rationalize experi
ences after they occur.
Sequence Effects. According to behavioral scientists, when people recall an experience, they don't remember
every single moment of it (unless the experience was
short and traumatic).lnstead, they recall a few significant
moments vividly and gloss over the others-they remem
ber snapshots, not movies. And they carry away an over-
Richard B. Chase is the justin Dart Professor cifOperations Management at the University of Southern California's
Marshall School of Business in Los Angeles. He is the author
cif"Where Does the Customer Fit in a Service Operation?"
(HBR November-December 1978) and "The Serv�ce Factory" (HBR july-August 1989, with David Garvin). Sri ram Dasu is an associate professor at the Marshall School.
80
all assessment of the experience that's based on three
factors: the trend in the sequence of pain or pleasure,
the high and low points, and the ending.
Not surprisingly, people prefer a sequence of experi
ences that improve over time. When gambling, they
prefer to lose $10 first, then win $5, rather than win $5,
then lose $10. There is also evidence that people pay
attention to the rate of improvement in a sequence
clearly preferring ones that improve faster. And, most
intriguing, the ending matters enormously. (See the side
bar "End on an Uptick:') A terrible ending usually domi
nates a person's recollection of an experience.
Duration Effects. Psychologists and cognitive scien tists have poured enormous effort into unraveling the
mysteries of how people process time. When do they pay
attention to the passage of time, and how do they esti
mate its duration? Although much of the mystery still
remains, one findjng that's been verified repeatedly is that
people who are mentally engaged in a task don't notice
how long it takes. Another is that, when prompted to pay
attention to the passage of time, people overestimate the
time elapsed. A third finding is that increasing the num
ber of segments in an encounter lengthens its perceived
duration. For example, a ten-minute dance sequence
consisting of four segments will seem longer than one
identical in length but split into two segments.
Since perceptions of time's passage are so subjective,
the obvious question is, When does duration matter?
Research indicates that unless an activity is much longer
or much shorter than expected, people pay little attention to its duration. There are two reasons for this. First, the
pleasurable content of the experience and how it is
arranged-rather than how long it takes-seem to domi
nate people's assessments. And second, aside from one-off transactions such as buying a cup of coffee, service
HARVARD BUSINESS REVIEW
encounters are rarely identical in length, so people have
only general reference points for evaluating duration.
Their estimates of how long it will take to visit a tax
accountant-or go to a ball game, or have minor surgery
are likely to be fuzzy.
Rationalization Effects. People desperately want things to make sense; if there's no handy explanation for
an unexpected event, they'll concoct one. Behavioral sci entists call this "counterfactual thinking;' but it's simpler
to call it second-guessing.
People second-guess because they want one clear rea
son for why something happened. In their mental simu
lation, they try to capture the specific what-ifs: "If only x
hadn't happened, things would be different!' Three characteristics stand out in this simulation. First, they
view the likely cause as a discrete thing, not a continuous,
intertwined process. For example, people are more likely
to blame a missed plane on "the backup in the tunnel"
than on a cluster of events that-in conjunction-caused
their late arrival. Second, people often conclude that de
viations from rituals and norms caused the unexpected
outcome. Professional sports are loaded with players who
follow rituals religiously: some baseball players avoid
stepping on the foul line at all costs, and many basket
ball players have particular dribbling routines before
shooting a free throw. Third, people tend to ascribe credit
or blame to individuals, not systems. Even when they
clearly see that the computer system caused the hotel
bill error, for example, they tend to blame the clerk.
They want to put a human face on the problem. One final
note about ascribing blame: people are far less apt to
"search for the guilty" if they think they've had some
control over the process that occurred. The more em powered and engaged they feel, the less angry they are
when something goes wrong.
Why Crui&e& Work
Modern cruise lines apply many of the operating
principles suggested by behavioral science.
Principle What Crui6e Line6 Do
Want to Perfect Your Company's Service?
In summary, people want explanations, and they'll
make them up if they have to. The explanation will nearly
always focus on something they can observe-something
that is discrete and concrete enough to be changed in
their if-only fantasies.
Several operating principles for service-encounter
management emerge from the behavioral-science find
ings we've just reviewed.
Principle 1 Finish Strong
Most service providers believe that the beginning and end
of an encounter-the so-called service bookends-are
equally weighted in the eyes of the customer. They're
dead wrong. The end is far more important because
it's what remains in the customer's recollections. Sure, it's
important to achieve a base level of satisfactory perfor
mance at the beginning, but a company is better off with
a relatively weak start and a modest upswing at the end
than with a boffo start and a so-so ending.
People's innate preference for improvement is another
factor in trus principle. We believe that the desire for im
provement applies not only to lengthy encounters but also to short, technology-mediated encounters, such as
on a Web site. The fact is, very few Web designers have
thought this issue through. Most companies spare no ex
pense to make their home pages attractive; a great deal
of thought goes into questions of aesthetics, content, and
navigation in the top page or two. This is an eminently
logical strategy, given the need to get people to enter and
engage with the site. However, too many Web encounters
start strong then go downhill fast. Our cursory review of commercial Web sites uncovered an alarming number
of problems: difficulty in exiting the site if an item is out of stock; difficulty in canceling an order if
the shipping charges are too high; no noti
fication of security for credit card informa
tion, and so on. Make no mistake, the frus
trated customer remembers the messy
final experience far more clearly than the
jazzy, supposedly sticky home page.
That which applies to short encounters
goes double for longer service encounters
Finish strong End each day on a high note with raffles, contests, shows, and so on.
like consulting projects. While it often
makes sense to pick low-hanging fruit
at the outset, a consultant would be well
advised (other things being equal) to
Segment the pleasure
Create rituals
JUNE 2001
End the cruise with the captain's dinner.
Pass out keepsakes or bottles of wine upon reaching home port.
Pack many events into one short vacation.
Offer captain's dinner and midnight buffets.
schedule the project so that a golden nugget or two appear at the end of the en
gagement. For instance, a consultant that's
hired to reengineer a company's business
processes might start with the distribution center and move to the call center later in
the project, because he knows from past
81
Want to Perfect Your Company's Serv1ce?
experience that the call center changes will likely reap a
windfall of savings. What you don't want is to have the
project results become less and less impressive, even if
(as is often the case) its labor costs are following a staged
decline. Even though one large consulting firm had performed admirably in its yearlong reengineering
project, for example, it received low marks from its client.
The consultants achieved more than the goals set, but the
lack of a visible upswing in results at the end left the im
pression of mediocrity. As it turns out, last impressions not first impressions-endure.
Compare that with another consulting project that
ended quite naturally on a high note. A statistician col league of ours was hired to determine what factors ac counted for the sales success of a new video game. The
client agreed at the start that the project would be a suc cess if the consultant's model could explain just 6% of the variability in sales among a dozen competing video
games. The consultant made progress over the first three months of the project, but it wasn't until the last day of the schedule that the analysis yielded a three-factor
combination that explained more than 90% of the vari
ability in sales. (For the garners out there, these factors
were kid testing, advertising, and the number of outlets
they could get the game into.) This positive surprise had
far more impact than it would have had at the outset,
since the clients' longer-term involvement had sensi tized them to the complexity of the task. Our col- league was lucky to deliver such a clear, better
than-anticipated result; he was luckier still
to have done so at the eleventh hour. Even if you can't end with a substan
tive bang, it's smart to finish with a styl istic flourish. Consider the airline in dustry, which suffers from high levels of customer dissatisfaction due to flight delays and cancellations, inadequate
legroom, and lost luggage. Without a doubt, those failures have to be ad dressed. But we'd guess that airlines could make up some ground if they paid
more attention to their customers' last
encounter- baggage collection. Why not offer a new service- aides to help
passengers in the baggage claim area? Simply having someone there would show concern for passengers.
Malaysian Airlines is one of the few carriers that understands that the encounter isn't over when the customer steps off the plane. Several years ago, an acquaintance was travel ing by Malaysian Airlines with her nine-month-old son. Even after nine years, she fondly recalls the help that the
82
flight attendants gave her with baggage collection and ground transportation. It cost the airline little to provide
that end-of-encounter assistance- and it gained a loyal
customer who's described that experience to fellow trav
elers dozens of times since. As simple as that example
sounds, such small touches have a disproportionate effect on customers' recollections.
Principle 2 Get the Bad Experiences Out of the Way Early
Behavioral science tells us that, in a sequence of events involving good and bad outcomes, people prefer to have undesirable events come first- so they can avoid dread
and to have desirable events come at the end of a se quence- so they can savor them.
This principle has concrete, immediate implications
for how health care professionals manage their encoun ters with patients. Imagine Danielle, a pediatric dental
hygienist, who has almost finished cleaning Asher's teeth. Asher, a skittish six-year-old and a frequent visitor to the
clinic, suffers from a mild form of gingivitis and has several cavities. Danielle accidentally scrapes a particu
larly sensitive spot, causing the boy momentary pain. She
still needs to clean two more teeth, which she is sure are
not as sensitive. She could either end the cleaning now and resume on the next visit, or she could complete it
today. Continuing would subject Asher to more discomfort, although it would be significantly
less than what he just felt. She is also wonder ing whether continuing (hence, increasing
the total pain) will affect Asher's percep
tion of the cleaning experience and his behavior on subsequent visits.
According to behavioral research, Danielle should finish the job. Asher
will carry away a better memory. He will remember the treatment, of course, but also that the pain "wasn't so bad at the end:' Danielle will have extended a painful expe
rience, yet because the ending was
slightly less painful, Asher's overall assessment of it will improve.
In any service encounter-from a simple pizza pickup to a complex, long-term consulting engagement perception is reality.
Most companies' services don't cause physical pain, ob viously. And often the discom fort that's part of a service en
counter occurs early naturally: the wait in line (unpleasant)
HARVARD BUSINESS REVIEW
comes before the meal or the theme park ride (pleasant).
When that's not the case, it may be necessary to extend
the encounter to soften the ending experience.
In professional services, the unpleasantness often comes in the form of bad news. Most people want bad
news brought to their attention right away. Unfortu nately, service providers are human just like the rest of us-they dread delivering bad news, so they delay it until
the last possible moment. This is exactly the wrong thing
to do. Get bad news, pain, discomfort, long waits in line, and other unpleasant things out of the way as soon as
possible so they don't dominate the customer's recollec tion of the entire experience.
Principle 3 Segment the Pleasure, Combine the Pain
As we noted earlier, experiences seem longer when they
are broken into segments. In addition, people have an asymmetric reaction to losses and gains. Compare win ning $10 in one gamble with winning $5 twice. Most of us
would prefer to win twice. What about losing $10 in one game as compared with losing $5 in each of two gambles?
Here, most people prefer only one loss. That's why com
panies should break pleasant experiences into multiple stages and combine unpleasant ones into a single stage.
Not many businesses have grasped this notion. Health care facilities, for example, typically make patients wait
at multiple points before they see the physician, but doing so makes the overall wait appear even longer. Clin ics would do better to let patients spend more time in the waiting room so they don't have to endure a second, third, or fourth wait in the examining rooms.
Phone help-line menus are frustrating in a similar way. To reach the department that can resolve a problem, a customer must listen to instructions and press (or voice)
a response. It often takes four or five such steps to get to the right place. Even if the actual time required to
run through, say, four menu queries is less than to run
through two, people recall four as taking longer. Service companies would do well to cut the number of steps it
takes to reach the final destination, thereby reducing the perceived pain of waiting.
The best trade shows have grasped both halves of this principle. They combine as many of the boring paper work steps as possible. The Internet World trade show, for example, lets attendees preregister over the Internet. When they arrive, they simply pick up a badge that's been programmed with personal data. The badges allow them to get information at any booth- attendees just swipe them through a reading device, thus avoiding an endless exchange of business cards and sign-in sheets. The things that attendees enjoy and come to see, such as product demos, are plentiful, and they're spread throughout the conference.
JUNE 2001
Want to Perfect Your Company's Service?
Disney's theme parks also understand both halves of
the principle. They do a great job of distracting customers
who are waiting in line, thus lessening their discomfort.
And they make the rides really short, as welJ. That's done
primarily so that more people can get on them, but this efficiency has the added benefit of segmenting the plea
sure, which in tum creates the perception of a longer and richer day at the theme park. From the customer's point of view, two 9o-second rides last longer than one three minute ride.
Principle 4 Build Commitment Through Choice
A fascinating study found that blood donors perceived significantly less discomfort when they were allowed to select the arm from which the blood would be drawn}
The lesson is clear: people are happier and more com
fortable when they believe they have some control over
a process, particularly an uncomfortable one. Often the
control handed over is largely symbolic (as in the choice of arm). In other cases, it's very real: the medical profes
sion has long recognized the value of allowing the patient to make an informed choice about alternative treatments for cancer and heart disease. These are
extremely important, high-stakes decisions, and great value is gained by including the patient in the decision.
He or she feels less helpless, less hopeless, and more committed to making the process work.
Many companies have learned to apply this principle in less life-threatening situations. Several airlines, for example, let passengers choose when they want to have their meal served during long flights. Most hotels give customers a choice of using an alarm clock or receiving a wake-up call. And some banks have moved away from snake line configurations and back to individual lines so
that customers can work with their favorite teller.
As one Midwestern company learned, this principle can both save money and make clients happy. Customers were complaining to the Xerox machine-servicing com
pany that repairs didn't happen quickly enough. At first,
the company considered adding more repair personnel, but upon reflection, it decided to give customers more
choice over the schedule. It let them determine the
urgency of the problem- service people would arrive faster for a critical failure than for a less urgent one. As expected, this improved customer satisfaction, but what surprised the company was that fewer repair people
were needed. The change also reduced the turnover of customer service reps because there were fewer schedul
ing conflicts with the customers. Conventional wisdom would say that allowing customers to pick the time would force the company to hire more staff. Here, however, as is often the case, customers actually wanted choice more than they wanted an instantaneous response.
83
Want to Perfect Your Company's Serv1ce?
Principle 5 Give People Rituals, and Stick to Them
The Right Remedy
Most service-encounter designers don't
realize just how ritualistic people are.
They find comfort, order, and meaning in
repetitive, familiar activities. Rituals are
particularly important in longer-term,
professional-service encounters: they're used to mark key moments in the rela
tionship, establish professional creden tials, create a feeling of inclusion, flatter
customers, set expectations, and get feed back. Common rituals include glowing introductions of staff at the start of an engagement, kickoff dinners, elegant
PowerPoint presentations, final celebra tions, and formal presentations to the
CEO (even though he or she may not have an interest in the project). Many rituals
are so small in scale that they're hard to
name. Nonetheless, they play an impor-
How do you make up for a service-encounter error? Research on what cus
tomers perceive as a fair remedy suggests that the answer depends on whether
1t IS an outcome error or a process error. A botched task calls for material com
pensation, while poor treatment from a server calls for an apology. Reversing
these recovery act1ons is unlikely to be effective.
Imagine being a copy store manager faced with two complaining customers.
One says that the job was done right but the clerk was surly. The other says that
the clerk was pleasant but when he got home, he realized that his report was
m1ssmg two pages, and he had to take 1t to a compet1tor near h1s house to get
the JOb done right. What should you do? In the case of the rude clerk, don't
give the customer some tangible compensation, such as a coupon for his next
visit. All the customer really wants is a sincere apology from the clerk and the
manager. In the case of the botched job, you can apologize all over the place, but
that won't satisfy the customer. He wants the job done right, and he wants some
compensation for his inconvenience. Thus, while apologies are appropriate m
both Situations, behavioral research clearly md1cates that process-based reme
dies should be applied to process-based problems and outcome-based remed1es
should be applied to outcome-based problems.
tant part in customers' perceptions of the experience. When McKinsey consultants listen to clients, for example,
they pepper pauses in the conversation with a character istic, noncommittal "uh-huh, uh-huh"that somebody once
labeled the "McKinsey grunt:• Sounds silly, but clients
notice when it's missing. Behavioral researchers have observed that these rituals
provide an implicit standard for evaluating service en
counters. Deviation from them is often cited as the cause
of a failure- particularly in professional services, where customers have difficulty evaluating precise causes and effects. Check in with customers after something's gone badly with a service engagement, and you'll find that this is quite true. "If Henry had covered the ten-step model on the new benefits system like Susan did for the old one, it wouldn't have flopped:'(The new system didn't require
the ten-step ritual; it failed for a constellation of other rea sons.) Or, "The consultant wearing the string tie was off in
his forecast by 10%�' (The dress code violation had nothing to do with the consultant's technical skills.)
It's easy to laugh at those examples, and more generally to dismiss people's tendency to focus on deviations from
norms and rituals when they're trying to explain a failure.
But make no mistake, behavioral science clearly shows
how critical rituals are in long-term relationships. Not getting the weekly call from the consultant on a project,
not copying the CEO on a progress report, not returning phone calls immediately- any of these lapses can be blamed after the fact for a failure. They can also, and even more ominously, shift a customer's perceptions about the
quality of the service, the service providers, and the com pany they represent.
84
Ultimately, only one thing really matters in a service
encounter-the customer's perceptions of what occurred.
Executives who design and oversee service encounters need to focus far more of their attention on the underly
ing factors affecting those perceptions. We believe that
service encounters can be engineered to enhance the customer's experience during the process and his or her
recollection of the process after it is completed. We've used science to explore those factors, but you'll
need to use your imagination to bring them alive. Put
yourself in your customers' shoes and imagine their jour ney. Visualize every moment they spend with you and your employees. Which of their encounters should be lengthened? Which should be shortened? Where in a
process are distractions most effective? Where should you offer choice to the customer? Which process rituals
should not be violated? What are the last images of your service that customers take away, and how can you en
hance them?
Behavioral science, applied with equal doses of empa thy and imagination, can improve service delivery. More
important, it can change the impressions that your cus
tomers remember, refer back to, and pass on to future customers. 1;7
1. R.T. Mills and D.S. Krantz, "lnfonnat1on, Cho1ce, and Reactions to Stress: A Field Experiment in a Blood Bank with Laboratory Analogue," jc>umal of Personality and Social Psychology, 1979.
Reprint R0106D To order reprints, see the last page of Exec utive Summaries.
To further explore the topic of this article, go to www.hbr.orglexplore.
HARVARD BUSINESS REVIEW
