Putting the Service-Profit

Chain to Work
by James L. Heskett, Thomas O. Jones, Gary W. Loveman,
W. Earl Sasser, Jr., and Leonard A. Schl esi nger
Repri nt 94204
Harvard Business Review
Copyri ght 1994 by the Presi dent and Fel l ows of Harvard Col l ege. Al l ri ghts reserved.
DRAWINGS BY GARISON WEILAND
Top-l evel executi ves of outstandi ng servi ce orga-
ni zati ons spend l i ttl e ti me setti ng profi t goal s or fo-
cusi ng on market share, the management mantra of
the 1970s and 1980s. Instead, they understand that
i n the new economi cs of servi ce, frontl i ne workers
and customers need to be the center of manage-
ment concern. Successful servi ce managers pay at-
tenti on to the factors that dri ve profi tabi l i ty i n thi s
new servi ce paradi gm: i nvestment i n peopl e, tech-
nol ogy that supports frontl i ne workers, revamped
recrui ti ng and trai ni ng practi ces, and compensa-
ti on l i nked to performance for empl oyees at every
l evel . And they express a vi si on of l eadershi p i n
terms rarel y heard i n corporate Ameri ca: an organi -
zati ons pati na of spi ri tual i ty, the i mportance of
the mundane.
A growi ng number of compani es that i ncl udes
Banc One, Intui t Corporati on, Southwest Ai rl i nes,
Servi ceMaster, USAA, Taco Bel l , and MCI know
that when they make empl oyees and customers
paramount, a radi cal shi ft occurs i n the way they
manage and measure success. The new economi cs
of servi ce requi res i nnovati ve measurement tech-
ni ques. These techni ques cal i brate the i mpact of
empl oyee sati sfacti on, l oyal ty, and producti vi ty on
the val ue of products and servi ces del i vered so that
managers can bui l d customer sati sfacti on and l oy-
al ty and assess the correspondi ng i mpact on prof-
i tabi l i ty and growth. In fact, the l i feti me val ue of
a l oyal customer can be astronomi cal , especi al l y
when referral s are added to the economi cs of cus-
tomer retenti on and repeat purchases of rel ated
products. For exampl e, the l i feti me revenue stream
from a l oyal pi zza eater can be $8,000, a Cadi l l ac
owner $332,000, and a corporate purchaser of com-
merci al ai rcraft l i teral l y bi l l i ons of dol l ars.
The servi ce-profi t chai n, devel oped from anal y-
ses of successful servi ce organi zati ons, puts hard
val ues on soft measures. It hel ps managers target
new i nvestments to devel op servi ce and sati sfac-
ti on l evel s for maxi mum competi ti ve i mpact, wi d-
eni ng the gap between servi ce l eaders and thei r
merel y good competi tors.
The Service-Profit Chain
The servi ce-profi t chai n establ i shes rel ati onshi ps
between profi tabi l i ty, customer l oyal ty, and em-
pl oyee sati sfacti on, l oyal ty, and producti vi ty. The
l i nks i n the chai n (whi ch shoul d be regarded as
proposi ti ons) are as fol l ows: Profi t and growth are
sti mul ated pri mari l y by customer l oyal ty. Loyal ty
When service companies put employees and customers first, a
radical shift occurs in the way they manage and measure success.
Putting the Service-Profit
Chain to Work
by James L. Heskett, Thomas O. Jones, Gary W. Loveman,
W. Earl Sasser, Jr., and Leonard A. Schl esi nger
James L. Heskett, Thomas O. Jones, Gary W. Loveman,
W. Earl Sasser, Jr., and Leonard A. Schlesinger are mem-
bers of the Harvard Business School faculty and service-
management interest group.
i s a di rect resul t of customer sati sfacti on. Sati sfac-
ti on i s l argel y i nfl uenced by the val ue of servi ces
provi ded to customers. Val ue i s created by sati sfi ed,
l oyal , and producti ve empl oyees. Empl oyee sati s-
facti on, i n turn, resul ts pri mari l y from hi gh-qual i ty
support servi ces and pol i ci es that enabl e empl oyees
to del i ver resul ts to customers. (See the chart, The
Li nks i n the Servi ce-Profi t Chai n. )
The servi ce-profi t chai n i s al so defi ned by a spe-
ci al ki nd of l eadershi p. CEOs of exempl ary servi ce
compani es emphasi ze the i mportance of each em-
pl oyee and customer. For these CEOs, the focus on
customers and empl oyees i s no empty sl ogan tai -
l ored to an annual management meeti ng. For exam-
pl e, Herbert Kel l eher, CEO of Southwest Ai rl i nes,
can be found aboard ai rpl anes, on tarmacs, and i n
termi nal s, i nteracti ng wi th empl oyees and cus-
tomers. Kel l eher bel i eves that hi ri ng empl oyees
that have the ri ght atti tude i s so i mportant that the
hi ri ng process takes on a pati na of spi ri tual i ty. In
addi ti on, he bel i eves that anyone who l ooks at
thi ngs sol el y i n terms of factors that can easi l y be
quanti fi ed i s mi ssi ng the heart of busi ness, whi ch i s
peopl e. Wi l l i am Pol l ard, the chai rman of Servi ce-
Master, conti nual l y underscores the i mportance of
teacher-l earner managers, who have what he
cal l s a servants heart. And John McCoy, CEO of
Banc One, stresses the uncommon partnershi p,
a system of support that provi des
max i mum l ati tude to i ndi v i dual
bank presi dents whi l e suppl yi ng i n-
for mati on sy stems and common
measurements of customer sati sfac-
ti on and fi nanci al measures.
A cl oser l ook at each l i nk reveal s
how the servi ce-profi t chai n func-
ti ons as a whol e.
Customer Loyalty Drives
Profitability and Growth
To maxi mi ze profi t, managers
have pursued the Hol y Grai l of
becomi ng number-one or -two i n
thei r i ndustri es for nearl y two
decades. Recentl y, however, new
measures of servi ce i ndustri es l i ke
software and banki ng suggest that
customer l oyal ty i s a more i mpor-
tan t deter mi n an t of pr ofi t. (See
Frederi ck F. Rei chhel d and W. Earl
Sasser, Jr., Zero Defecti ons: Qual i ty
Comes to Servi ces, HBR September-
October 1990.) Rei chhel d and Sasser
esti mate that a 5% i ncrease i n cus-
tomer l oyal ty can produce profi t i ncreases from
25% to 85%. They concl ude that quality of market
share, measured i n terms of customer l oyal ty, de-
serves as much attenti on as quantity of share.
Banc One, based i n Col umbus, Ohi o, has devel -
oped a sophi sti cated system to track several factors
i nvol ved i n customer l oyal ty and sati sfacti on. Once
dri ven stri ctl y by fi nanci al measures, Banc One
now conducts quarterl y measures of customer re-
tenti on; the number of servi ces used by each cus-
tomer, or depth of relationship; and the l evel of
customer sati sfacti on. The strategi es deri ved from
thi s i nformati on hel p expl ai n why Banc One has
achi eved a return on assets more than doubl e that
of i ts competi tors i n recent years.
Customer Satisfaction Drives
Customer Loyalty
Leadi ng servi ce compani es are currentl y tryi ng to
quanti fy customer sati sfacti on. For exampl e, for
several years, Xerox has pol l ed 480,000 customers
per year regardi ng product and servi ce sati sfacti on
usi ng a fi ve-poi nt scal e from 5 (hi gh) to 1 (l ow). Un-
ti l two years ago, Xeroxs goal was to achi eve 100%
4s (sati sfi ed) and 5s (very sati sfi ed) by the end of
1993. But i n 1991, an anal ysi s of customers who
gave Xerox 4s and 5s on sati sfacti on found that the
The lifetime value of a loyal pizza eater can be as much as $8,000.
HARVARD BUSINESS REVIEW March-Apri l 1994 165
166 HARVARD BUSINESS REVIEW March-Apri l 1994
rel ati onshi ps between the scores and actual l oyal ty
di ffered greatl y dependi ng on whether the cus-
tomers were very sati sfi ed or sati sfi ed. Customers
gi vi ng Xerox 5s were si x ti mes more l i kel y to repur-
chase Xerox equi pment than those gi vi ng 4s.
Thi s anal ysi s l ed Xerox to extend i ts efforts to
create apostlesa term coi ned by Scott D. Cook,
CEO of software producer and di stri butor, Intui t
Corporati on, descri bi ng customers so sati sfi ed that
they convert the uni ni ti ated to a product or servi ce.
Xeroxs management currentl y wants to achi eve
100% apostl es, or 5s, by the end of 1996 by up-
gradi ng servi ce l evel s and guaranteei ng customer
sati sfacti on. But just as i mportant for Xerox s
profi tabi l i ty i s to avoi d creati ng terrorists: cus-
tomers so unhappy that they speak out agai nst a
poorl y del i vered servi ce at every opportuni ty. Ter-
rori sts can reach hundreds of potenti al customers.
I n some i nstances, they can even di scourage ac-
quai ntances from tryi ng a servi ce or product. (See
the graph A Sati sfi ed Customer Is Loyal . )
Value Drives Customer Satisfaction
Customers today are strongl y val ue ori ented. But
just what does that mean? Customers tel l us that
val ue means the resul ts they recei ve i n rel ati on to
the total costs (both the pri ce and other costs to cus-
tomers i ncurred i n acqui ri ng the servi ce). The i n-
surance company, Progressi ve Corporati on, i s cre-
ati ng just thi s ki nd of val ue for i ts customers by
processi ng and payi ng cl ai ms qui ckl y and wi th l i t-
tl e pol i cyhol der effort. Members of the companys
CAT (catastrophe) team fl y to the scene of major ac-
ci dents, provi di ng support servi ces l i ke transporta-
ti on and housi ng and handl i ng cl ai ms rapi dl y. By re-
duci ng l egal costs and actual l y pl aci ng more money
i n the hands of the i njured parti es, the CAT team
more than makes up for the added expenses the or-
gani zati on i ncurs by mai ntai ni ng the team. In addi -
ti on, the CAT team del i vers val ue to customers,
whi ch hel ps expl ai n why Progressi ve has one of the
hi ghest margi ns i n the property-and-casual ty i nsur-
ance i ndustry.
Employee Productivity Drives Value
At Southwest Ai rl i nes, the seventh-l argest U.S.
domesti c carri er, an astoni shi ng story of empl oy-
ee producti vi ty occurs dai l y. Ei ghty-si x percent of
the companys 14,000 empl oyees are uni oni zed. Po-
si ti ons are desi gned so that empl oyees can per-
form several jobs i f necessary. Schedul es, routes,
and company practi ces such as open seati ng and
The Links in the Service-Profit Chain
workplace design
job design
employee selection
and development
employee rewards
and recognition
tools for serving customers
service designed and
delivered to meet
targeted customers needs
retention
repeat business
referral
service concept:
results for customers
Operating Strategy and
Service Delivery System
Internal
Service
Quality
Employee
Satisfaction
Employee
Retention
Employee
Productivity
External
Service
Value
Customer
Satisfaction
Customer
Loyalty
Revenue
Growth
Profitablity
HARVARD BUSINESS REVIEW March-Apri l 1994 167
THE SERVICE-PROFIT CHAIN
the use of si mpl e, col or-coded, reusabl e boardi ng
passes enabl e the boardi ng of three and four ti mes
more passengers per day than competi ng ai rl i nes. In
fact, Southwest depl anes and rel oads two-thi rds of
i ts fl i ghts i n 15 mi nutes or l ess. Because of ai rcraft
avai l abi l i ty and short-haul routes that dont requi re
l ong l ayovers for fl i ght crews, Southwest has rough-
l y 40% more pi l ot and ai rcraft uti l i zati on than i ts
major competi tors: i ts pi l ots fl y on average 70 hours
per month versus 50 hours at other ai rl i nes. These
factors expl ai n how the company can charge fares
from 60% to 70% l ower than exi sti ng fares i n mar-
kets i t enters.
At Southwest, customer percepti ons of val ue are
very hi gh, even though the ai rl i ne does not assi gn
seats, offer meal s, or i ntegrate i ts reservati on sys-
tem wi th other ai rl i nes. Customers pl ace hi gh val -
ue on Southwests frequent departures, on-ti me
servi ce, fri endl y empl oyees, and very l ow fares.
Southwests management knows thi s because i ts
major marketi ng research uni t i ts 14,000 empl oy-
ees i s i n dai l y contact wi th customers and reports
i ts fi ndi ngs back to management. In addi ti on, the
Federal Avi ati on Admi ni strati ons performance
measures show that Southwest, of al l the major ai r-
l i nes, regul arl y achi eves the hi ghest l evel of on-
ti me arri val s, the l owest number of compl ai nts,
and the fewest l ost-baggage cl ai ms per 1,000 pas-
sengers. When combi ned wi th
Southwests l ow fares per seat-
mi l e, these i ndi cators show the
hi gher val ue del i vered by South-
wests empl oyees compared wi th
mos t domes ti c competi tor s .
Southwest has been profi tabl e for
21 consecuti ve years and was the
onl y major ai rl i ne to real i ze a
pr ofi t i n 1992. (See th e gr aph
How Southwest Compares wi th
Its Competi tors. )
Employee Loyalty Drives
Productivity
Tradi ti onal measur es of the
l os s es i n cu r r ed by empl oy ee
turnover concentrate onl y on the
cost of recr ui ti ng, hi ri ng, and
trai ni ng repl acements. I n most
s er v i ce j obs , th e r eal cos t of
turnover i s the l oss of producti vi -
ty and decreased customer sati s-
facti on. One recent study of an
automobi l e deal ers sal es person-
nel by Abt Associ ates concl uded
that the average monthl y cost of repl aci ng a sal es
representati ve who had fi ve to ei ght years of experi -
ence wi th an empl oyee who had l ess than one year
of experi ence was as much as $36,000 i n sal es. And
the costs of l osi ng a val ued broker at a securi ti es
fi rm can be sti l l more di re. Conservati vel y esti mat-
ed, i t takes nearl y fi ve years for a broker to rebui l d
rel ati onshi ps wi th customers that can return $1
mi l l i on per year i n commi ssi ons to the brokerage
house a cumul ati ve l oss of at l east $2.5 mi l l i on i n
commi ssi ons.
Employee Satisfaction Drives Loyalty
In one 1991 propri etary study of a property-and-
casual ty i nsurance companys empl oyees, 30% of
al l di ssati sfi ed empl oyees regi stered an i ntenti on to
l eave the company, a potenti al turnover rate three
ti mes hi gher than that for sati sfi ed empl oyees. In
thi s same case, l ow empl oyee turnover was found
to be l i nked cl osel y to hi gh customer sati sfacti on.
I n contrast, Southwest Ai rl i nes, recentl y named
one of the countrys ten best pl aces to work, experi -
ences the hi ghest rate of empl oyee retenti on i n the
ai rl i ne i ndustry. Sati sfacti on l evel s are so hi gh that
at some of i ts oper ati n g l ocati on s, empl oy ee
turnover rates are l ess than 5% per year. USAA, a
major provi der of i nsurance and other fi nanci al ser-
A Satisfied Customer Is Loyal
100%
80
60
40
20
terrorist
L
o
y
a
l
t
y
(
R
e
t
e
n
t
i
o
n
)
1
extremely
dissatisfied
2
somewhat
dissatisfied
3
slightly
dissatisfied
4
satisfied
5
very
satisfied
Satisfaction Measure
apostle
zone of defection
zone of indifference
zone of affection
THE SERVICE-PROFIT CHAIN
168 HARVARD BUSINESS REVIEW March-Apri l 1994
vi ces by di rect mai l and phone, al so achi eves l ow
l evel s of empl oyee turnover by ensuri ng that i ts
empl oyees are hi ghl y sati sfi ed. But what dri ves em-
pl oyee sati sfacti on? Is i t compensati on, perks, or
pl ush workpl aces?
Internal Quality Drives
Employee Satisfaction
What we cal l the internal quality of a worki ng
envi ronment contri butes most to empl oyee sati s-
facti on. Internal qual i ty i s measured by the feel i ngs
that empl oyees have toward thei r jobs, col l eagues,
and compani es. What do servi ce empl oyees val ue
most on the job? Al though our data are prel i mi nary
at best, they poi nt i ncreasi ngl y to the abi l i ty and
authori ty of servi ce workers to achi eve resul ts for
customers. At USAA, for exampl e, tel ephone sal es
and servi ce representati ves are backed by a sophi s-
ti cated i nformati on system that puts compl ete cus-
tomer i nformati on fi l es at thei r fi ngerti ps the i n-
stant they recei ve a customers cal l . In addi ti on,
state-of-the-art, job-rel ated trai ni ng i s made avai l -
abl e to USAA empl oyees. And the curri cul um goes
sti l l further, wi th 200 courses i n 75 cl assrooms on
a wi de range of subjects.
Internal qual i ty i s al so characteri zed by the atti -
tudes that peopl e have toward one another and the
way peopl e serve each other i nsi de the organi za-
ti on. For exampl e, Servi ceMaster, a provi der of a
range of cl eani ng and mai ntenance servi ces, ai ms
to maxi mi ze the di gni ty of the i ndi vi dual servi ce
worker. Each year, i t anal yzes i n depth a part of the
mai ntenance process, such as cl eani ng a fl oor, i n or-
der to reduce the ti me and effort needed to com-
pl ete the task. The i mportance of the mundane i s
stressed repeatedl y i n Servi ceMasters manage-
ment trai ni ngfor exampl e, i n the seven-step pro-
cess devi sed for cl eani ng a hospi tal room: from the
fi rst step, greeti ng the pati ent, to the l ast step, ask-
i ng pati ents whether or not they need anythi ng el se
done. Usi ng thi s process, servi ce workers devel op
communi cati on ski l l s and l earn to i nteract wi th
pati ents i n ways that add depth and di mensi on
to thei r jobs.
Leadership Underlies the
Chains Success
Leaders who understand the servi ce-profi t chai n
devel op and mai ntai n a corporate cul ture centered
around servi ce to customers and fel l ow empl oyees.
They di spl ay a wi l l i ngness and abi l i ty to l i sten.
Successful CEOs l i k e John Marti n of Taco Bel l ,
John McCoy of Banc One, Herb Kel l eher of South-
west, and Bi l l Pol l ard of Servi ceMaster spend a great
deal of ti me wi th customers and empl oyees, experi -
enci ng thei r compani es servi ce processes whi l e l i s-
teni ng to empl oyees for suggesti ons for i mprove-
ment. They care about thei r empl oyees and spend
a great deal of ti me sel ecti ng, tracki ng, and rec-
ogni zi ng them.
For exampl e, Bri gadi er General Robert McDer-
mott, unti l recentl y chai rman and CEO of USAA,
refl ected, Publ i c recogni ti on of outstandi ng em-
pl oyees fl ows natural l y from our corporate cul ture.
That cul ture i s tal ked about al l the ti me, and we
l i ve i t. Accordi ng to Scott Cook at Intui t, Most
peopl e take cul ture as a gi ven. It i s around you, the
thi nki ng goes, and you cant do anythi ng about i t.
How Southwest Compares with Its Competitors
Revenue
in billions
Profit
in millions
On-time
arrival
A
m
e
r
i
c
a
n
D
e
l
t
a
N
o
r
t
h
w
e
s
t
U
n
i
t
e
d
U
S
A
i
r
S
o
u
t
h
w
e
s
t
A
m
e
r
i
c
a
n
D
e
l
t
a
N
o
r
t
h
w
e
s
t
U
n
i
t
e
d
U
S
A
i
r
S
o
u
t
h
w
e
s
t
A
m
e
r
i
c
a
n
D
e
l
t
a
N
o
r
t
h
w
e
s
t
U
n
i
t
e
d
U
S
A
i
r
S
o
u
t
h
w
e
s
t
0
2
4
6
8
10
12
$14
-600
-500
-400
-300
-200
-100
0
$100
80
82
84
86
88
90
92
94%
78
HARVARD BUSINESS REVIEW March-Apri l 1994 169
However, when you run a company, you have the
opportuni ty to determi ne the cul ture. I fi nd that
when you champi on the most nobl e val ues i n-
cl udi ng servi ce, anal ysi s, and database deci si on
maki ngempl oyees ri se to the chal l enge, and you
forever change thei r l i ves.
Relating Links in the Chain for
Management Action
Whi l e many organi zati ons are begi nni ng to mea-
sure rel ati onshi ps between i ndi vi dual l i nks i n the
servi ce-profi t chai n, onl y a few have rel ated the
l i nks i n meani ngful ways ways that can l ead to
comprehensi v e strategi es for achi ev i ng l asti ng
competi ti ve advantage.
The 1991 propri etary study of a property-and-ca-
sual ty i nsurance company, ci ted earl i er, not onl y
i denti fi ed the l i nks between empl oyee sati sfacti on
and l oyal ty but al so establ i shed that a pri mary
source of job sati sfacti on was the servi ce workers
percepti ons of thei r abi l i ty to meet customer needs.
Those who fel t they di d meet customer needs regi s-
tered job sati sfacti on l evel s more than twi ce as hi gh
as those who fel t they di dnt. But even more i mpor-
tant, the same study found that when a servi ce
worker l eft the company, customer sati sfacti on l ev-
el s dropped sharpl y from 75% to 55%. As a resul t
of thi s anal ysi s, management i s tryi ng to reduce
turnover among customer-contact empl oyees and
to enhance thei r job ski l l s.
Si mi l arl y, i n a study of i ts seven tel ephone cus-
tomer servi ce centers, MCI found cl ear rel ati on-
shi ps between empl oyees percepti ons of the qual i -
ty of MCI servi ce and empl oyee sati sfacti on. The
study al so l i nked empl oyee sati sfacti on di rectl y to
customer sati sfacti on and i ntenti ons to conti nue to
use MCI servi ces. Identi fyi ng these rel ati onshi ps
moti vated MCIs management to probe deeper and
determi ne what affected job sati sfacti on at the ser-
vi ce centers. The factors they uncovered, i n order of
i mportance, were sati sfacti on wi th the job i tsel f,
trai ni ng, pay, advancement fai rness, treatment
wi th respect and di gni ty, teamwork, and the com-
panys i nterest i n empl oyees wel l -bei ng. Armed
wi th thi s i nformati on, MCIs management began
exami ni ng i ts pol i ci es concerni ng those i tems val -
ued most by empl oyees at i ts servi ce centers. MCI
has i ncorporated i nformati on about i ts servi ce ca-
pabi l i ti es i nto trai ni ng and communi cati ons efforts
and tel evi si on adverti si ng.
No organi zati on has made a more comprehensi ve
effort to measure rel ati onshi ps i n the servi ce-profi t
chai n and fashi on a strategy around them than the
fast-food company, Taco Bel l , a subsi di ary of Pepsi -
Co. Taco Bel l s management tracks profi ts dai l y by
uni t, market manager, zone, and country. By i n-
tegrati ng thi s i nformati on wi th the resul ts of exi t
i ntervi ews that Taco Bel l conducts wi th 800,000
customers annual l y, management has found that
stores i n the top quadrant of customer sati sfacti on
rati ngs outperform the others by al l measures. As a
resul t, i t has l i nked no l ess than 20% of al l opera-
ti ons managers compensati on i n company-owned
stores to customer sati sfacti on rati ngs, real i zi ng a
subsequent i ncrease i n both customer sati sfacti on
rati ngs and profi ts.
However, Taco Bel l s efforts dont stop there. By
exami ni ng empl oyee turnover records for i ndi vi du-
al stores, Taco Bel l has di scovered that the 20% of
Consumer complaints
per 1,000 passengers
Passengers
per employee
Employees
per aircraft
A
m
e
r
i
c
a
n
D
e
l
t
a
N
o
r
t
h
w
e
s
t
U
n
i
t
e
d
U
S
A
i
r
S
o
u
t
h
w
e
s
t
A
m
e
r
i
c
a
n
D
e
l
t
a
N
o
r
t
h
w
e
s
t
U
n
i
t
e
d
U
S
A
i
r
S
o
u
t
h
w
e
s
t
A
m
e
r
i
c
a
n
D
e
l
t
a
N
o
r
t
h
w
e
s
t
U
n
i
t
e
d
U
S
A
i
r
S
o
u
t
h
w
e
s
t
0
.2
.4
.6
.8
1.0
1.2
1.4
0
500
1,000
1,500
2,000
2,500
0
20
40
60
80
100
120
140
160
THE SERVICE-PROFIT CHAIN
170 HARVARD BUSINESS REVIEW March-Apri l 1994
the stores wi th the l owest turnover rates enjoy dou-
bl e the sal es and 55% hi gher profi ts than the 20% of
stores wi th the hi ghest empl oyee turnover rates. As
a resul t of thi s sel f-exami nati on, Taco Bel l has i n-
sti tuted fi nanci al and other i ncenti ves i n order to
reverse the cycl e of fai l ure that i s associ ated wi th
poor empl oyee sel ecti on, subpar trai ni ng, l ow pay,
and hi gh turnover.
In addi ti on, Taco Bel l moni tors i nternal qual i ty
through a network of 800 numbers created to an-
swer empl oyees questi ons, fi el d thei r compl ai nts,
remedy si tuati ons, and al ert top-l evel management
to potenti al troubl e spots. It al so conducts peri odi c
empl oyee roundtabl e meeti ngs, i ntervi ews, as wel l
as a comprehensi ve companywi de survey every
two or three years i n order to measure sati sfacti on.
As a resul t of al l thi s work, Taco Bel l s empl oyee
sati sfacti on program features a new sel ecti on pro-
cess, i mproved ski l l bui l di ng, i ncreased l ati tude for
deci si on maki ng on the job, further automati on of
unpl easant back room l abor.
Rel ati ng al l the l i nks i n the servi ce-profi t chai n
may seem to be a tal l order. But profi tabi l i ty de-
pends not onl y on pl aci ng hard val ues on soft mea-
sures but al so on l i nki ng those i ndi vi dual measures
together i nto a comprehensi ve servi ce pi cture. Ser-
vi ce organi zati ons need to quanti fy thei r i nvest-
ments i n peopl e both customers and empl oyees.
The servi ce-profi t chai n provi des the framework for
thi s cri ti cal task.
Service-Profit Chain Audit
A servi ce-profi t chai n audi t hel ps compani es de-
termi ne what dri ves thei r profi t and suggests ac-
ti ons that can l ead to l ong-term profi tabi l i ty. As
they revi ew the audi t, managers shoul d ask them-
sel ves what efforts are under way to obtai n answers
to the fol l owi ng questi ons and what those answers
reveal about thei r compani es.
Profit and Growth
1. How do we define loyal customers?
Customers often become more profi tabl e over
ti me. And l oyal customers account for an unusual -
l y hi gh proporti on of the sal es and profi t growth of
successful servi ce provi ders. In some organi zati ons,
l oyal ty i s measured i n terms of whether or not a
customer i s on the company rol l s. But several com-
pani es have found that thei r most l oyal customers
the top 20% of total customers not onl y provi de
al l the profi t but al so cover l osses i ncurred i n deal -
i ng wi th l ess l oyal customers.
Because of the l i nk between l oyal customers and
profi t, Banc One measures depth of relationship
the number of avai l abl e rel ated fi nanci al servi ces,
such as checki ng, l endi ng, and safe deposi t, actual -
l y used by customers. Recogni zi ng the same re-
l ati onshi p, Taco Bel l measures share of stomach
to assess the companys sal es agai nst al l other
food purchases a customer can potenti al l y make. As
a resul t, the fast-food chai n i s try i ng to reach
consumers through ki osks, carts, trucks, and the
shel ves of supermarkets.
2. Do measurements of customer profitability in-
clude profits from referrals?
Compani es that measure the stream of revenue
and profi ts from l oyal customers (retenti on) and re-
peat sal es often overl ook what can be the most i m-
portant of the three Rs of l oyal ty: referral s. For ex-
ampl e, Intui t provi des hi gh-qual i ty, free l i feti me
servi ce for a personal fi nance software package that
sel l s for as l i ttl e as $30. The strategy makes sense
HARVARD BUSINESS REVIEW March-Apri l 1994 171
when the val ue of a l oyal customer i s consi dered a
revenue stream of several thousands of dol l ars from
software updates, suppl i es, and new customer refer-
ral s. Wi th thi s strategy i n pl ace, Intui t i ncreased i ts
sal es to more than $30 mi l l i on wi th just two U.S.
fi el d sal es representati ves.
3. What proportion of business development ex-
penditures and incentives are directed to the reten-
tion of existing customers?
Too many compani es concentrate nearl y al l thei r
efforts on attracti ng new customers. But i n busi -
nesses l i ke l i fe i nsurance, a new pol i cyhol der does-
nt become profi tabl e for at l east three years. In the
credi t-card fi nance busi ness, the break-even poi nt
for a new customer i s often si x or more years be-
cause of hi gh-marketi ng and bad-debt costs i n the
fi rst year of a rel ati onshi p wi th cardhol ders. These
costs must be defrayed by profi ts from l oyal cus-
tomers, suggesti ng the need for a careful di vi si on of
organi zati onal effort between customer retenti on
and devel opment.
4. Why do our customers defect?
Its i mportant to fi nd out not onl y where defec-
tors go but al so why they defect. Was i t because of
poor servi ce, pri ce, or val ue? Answers to these ques-
ti ons provi de i nformati on about whether or not ex-
i sti ng strategi es are worki ng. In addi ti on, exi t i nter-
vi ews of customers can have real sal es i mpact. For
exampl e, at one credi t-card servi ce organi zati on, a
phone cal l to questi on cardhol ders who had stopped
usi ng thei r cards l ed to the i mmedi ate rei nstate-
ment of one-thi rd of the defectors.
Customer Satisfaction
5. Are customer satisfaction data gathered in an
objective, consistent, and periodic fashion?
Currentl y, the weak est measurements bei ng
used by the compani es we have studi ed concern
customer sati sfacti on. At some compani es, hi gh
l evel s of reported customer sati sfacti on are contra-
di cted by conti nui ng decl i nes i n sal es and profi ts.
Upon cl oser observati on, we di scovered that the
servi ce provi ders were gami ng the data, usi ng
mani pul ati ve methods for col l ecti ng customer sat-
i sfacti on data. In one extreme case, an automobi l e
deal er sent a questi onnai re to recent buyers wi th
the hi ghest marks al ready fi l l ed i n, requi ri ng own-
ers to al ter the marks onl y i f they di sagreed. Com-
pani es can, however, obtai n more objecti ve resul ts
usi ng thi rd party i ntervi ews; mystery shop-
pi ng by uni denti fi ed, pai d observers; or technol o-
gi es l i ke touch-screen tel evi si on.
Consi stency i s at l east as i mportant as the actual
questi ons asked of customers. Some of Banc Ones
operati ng uni ts formerl y conducted thei r own cus-
tomer sati sfacti on surveys. Today the surveys have
been central i zed, made mandatory, and are admi n-
i stered by mai l on a quarterl y basi s to around
125,000 customers. When combi ned wi th peri odi c
measurement, the surveys provi de hi ghl y rel evant
trend i nformati on that i nforms the manageri al de-
ci si on-maki ng process. Si mi l arl y, Xeroxs measures
of sati sfacti on obtai ned from 10,000 customers per
month a product of an unchangi ng set of survey
questi ons and very l arge sampl es make possi bl e
peri od-to-peri od compari sons that are i mportant i n
measuri ng and rewardi ng performance.
6. Where are the listening posts for obtaining cus-
tomer feedback in your organization?
Li steni ng posts are tool s for col l ecti ng data from
customers and systemati cal l y transl ati ng those da-
ta i nto i nformati on i n order to i mprove servi ce and
products. Common exampl es are l etters of com-
pl ai nt. Sti l l more i mportant l i steni ng posts are re-
ports from fi el d sal es and servi ce personnel or the
l ogs of tel ephone servi ce representati ves. Intui ts
content anal ysi s of customer servi ce i nqui ri es fi el d-
ed by servi ce representati ves produced over 50 soft-
ware i mprovements and 100 software documenta-
ti on i mprovements i n a si ngl e year. USAA has gone
one step further by automati ng the feedback pro-
cess to enter data onl i ne, enabl i ng i ts anal ysi s and
pl ans departments to devel op correcti ve acti ons.
7. How is information concerning customer satis-
faction used to solve customer problems?
I n order to handl e customer probl ems, servi ce
provi ders must have the l ati tude to resol ve any si t-
uati on promptl y. In addi ti on, i nformati on regardi ng
a customer concern must be transmi tted to the ser-
vi ce provi der qui ckl y. Customers and empl oyees
must be encouraged to report rather than suppress
concerns. For exampl e, one Boston-area Lexus deal -
er noti fi ed i ts customers, If you are experi enci ng a
probl em wi th your car or our servi ce department
and you cant answer 100% sati sfi ed when you re-
cei ve your survey di rectl y from Lexus, pl ease gi ve
us the opportuni ty to correct the probl em before
you fi l l out the survey. Lexus takes i ts customer
surveys very seri ousl y.
External Service Value
8. How do you measure service value?
Val ue i s a functi on not onl y of costs to the cus-
tomer but al so of the resul ts achi eved for the cus-
THE SERVICE-PROFIT CHAIN
172 HARVARD BUSINESS REVIEW March-Apri l 1994
tomer. Val ue i s al ways rel ati ve because i t i s based
both on percepti ons of the way a servi ce i s del i vered
and on i ni ti al customer expectati ons. Typi cal l y, a
company measures val ue usi ng the reasons ex-
pressed by customers for hi gh or l ow sati sfacti on.
Because val ue vari es wi th i ndi vi dual expectati ons,
efforts to i mprove val ue i nevi tabl y requi re servi ce
organi zati ons to move al l l evel s of management
cl oser to the customer and gi ve frontl i ne servi ce
empl oyees the l ati tude to customi ze a standard ser-
vi ce to i ndi vi dual needs.
9. How is information concerning customers per-
ceptions of value shared with those responsible for
designing a product or service?
Rel ayi ng i nformati on concerni ng customer ex-
pectati ons to those responsi bl e for desi gn often re-
qui res the formati on of teams of peopl e responsi bl e
for sal es, operati ons, and servi ce or product desi gn,
as wel l as the frequent assi gnment of servi ce de-
si gners to tasks requi ri ng fi el d contact wi th cus-
tomers. Intui t has created thi s ki nd of capabi l i ty i n
product devel opment teams. And al l Intui t empl oy-
ees, i ncl udi ng the CEO, must peri odi cal l y work on
the customer servi ce phones. Si mi l arl y, at South-
west, those responsi bl e for fl i ght schedul i ng peri od-
i cal l y work shi fts i n the companys termi nal s to get
a feel for the i mpact of schedul es on customer and
empl oyee sati sfacti on.
10. To what extent are measures taken of differ-
ences between customers perceptions of quality
delivered and their expectations before delivery?
Ul ti matel y, servi ce qual i ty i s a functi on of the
gap between percepti ons of the actual servi ce expe-
ri enced and what a customer expected before re-
cei vi ng that servi ce. Actual servi ce i ncl udes both
fi nal resul ts and the process through whi ch those
resul ts were obtai ned. Di fferences between experi -
ences and expectati ons can be measured i n generi c
di mensi ons such as the rel i abi l i ty and ti mel i ness of
servi ce, the empathy and authori ty wi th whi ch the
servi ce was del i vered, and the extent to whi ch the
customer i s l eft wi th tangi bl e evi dence (l i ke a cal l -
i ng card) that the servi ce has been performed.
11. Do our organizations efforts to improve exter-
nal service quality emphasize effective recovery
from service errors in addition to providing a ser-
vice right the first time?
A popul ar concept of qual i ty i n manufacturi ng i s
the i mportance of doi ng thi ngs ri ght the fi rst
ti me. But customers of servi ce organi zati ons often
al l ow one mi stake. Some organi zati ons are very
At Southwest Airlines, frequent fliers participate in the auditioning and selection of cabin attendants.
HARVARD BUSINESS REVIEW March-Apri l 1994 173
good at del i veri ng servi ce as l ong as nothi ng goes
wrong. Others organi ze for and thri ve on servi ce
emergenci es. Outstandi ng servi ce organi zati ons do
both by gi vi ng frontl i ne empl oyees the l ati tude to
effect recovery. Southwest Ai rl i nes mai ntai ns a
pol i cy of al l owi ng frontl i ne empl oyees to do what-
ever they feel comfortabl e doi ng i n order to sati sfy
customers. Xerox authori zes frontl i ne servi ce em-
pl oyees to repl ace up to $250,000 worth of equi p-
ment i f customers are not getti ng resul ts.
Employee Productivity
12. How do you measure employee productivity?
13. To what extent do measures of productivity
identify changes in the quality as well as the quan-
tity of service produced per unit of input?
In many servi ces, the ul ti mate measure of qual i -
ty may be customer sati sfacti on. That measure
shoul d be combi ned wi th measures of quanti ty to
determi ne the total output of the servi ce organi za-
ti on. At Servi ceMaster, for exampl e, measures of
output i n the school s and hospi tal s cl eaned under
the companys supervi si on i ncl ude both numbers of
work orders performed per empl oyee hour and the
qual i ty of the work done, as determi ned by peri odi c
i nspecti ons performed by Servi ceMaster and cl i -
ent personnel . Si mi l arl y, Southwest Ai rl i nes del i v-
ers rel ati vel y hi gh l evel s of producti vi ty i n terms
of both qual i ty and quanti ty. In fact, outstandi ng
ser v i ce competi tor s ar e r epl aci ng the ty pi cal
ei ther/or trade-off between qual i ty and quanti ty
wi th an and/al so i mperati ve.
Employee Loyalty
14. How do you create employee loyalty?
Empl oyee l oyal ty goes hand i n hand wi th produc-
ti vi ty, contradi cti ng the conventi onal wi sdom that
successful servi ce provi ders shoul d be promoted to
l arger supervi sory responsi bi l i ti es or moved to a
si mi l ar job i n a l arger busi ness uni t. Servi ceMaster
and Taco Bel l have expanded jobs wi thout promot-
i ng good servi ce work ers away from thei r cus-
tomers. At Servi ceMaster, effecti ve si ngl e-uni t
managers are gi ven supervi sory responsi bi l i ti es for
custodi al , mai ntenance, or other workers at more
than one hospi tal or school . Taco Bel l gi ves restau-
rant general managers a hunti ng l i cense to devel -
op new sal es si tes i n the nei ghborhoods served by
thei r restaurants and rewards them for doi ng i t.
15. Have we made an effort to determine the right
level of employee retention?
Rarel y i s the ri ght l evel of retenti on 100%. Dy-
nami c servi ce organi zati ons requi re a certai n l evel
of turnover. However, i n cal i brati ng desi red turn-
over l evel s, i t i s i mportant to take i nto account the
ful l cost of the l oss of key servi ce provi ders, i ncl ud-
i ng those of l ost sal es and producti vi ty and added
recrui ti ng, sel ecti on, and trai ni ng.
Employee Satisfaction
16. I s employee satisfaction measured in ways that
can be linked to similar measures of customer sat-
isfaction with sufficient frequency and consisten-
cy to establish trends for management use?
Taco Bel l studi es empl oyee sati sfacti on through
surveys, frequent i ntervi ews, and roundtabl e meet-
i ngs. Customer sati sfacti on i s measured by i nter-
vi ews wi th customers conducted bi annual l y and
i ncl udes questi ons about sati sfacti on wi th empl oy-
ee fri endl i ness and hustl e. Both the empl oyee and
customer sati sfacti on ranki ngs are comprehensi ve,
store-speci fi c, and conducted frequentl y. Wi th
these data, the company can better understand
overal l trends and the l i nks between empl oyee and
customer sati sfacti on.
17. Are employee selection criteria and methods
geared to what customers, as well as managers, be-
lieve are important?
At Southwest Ai rl i nes, for exampl e, frequent
fl i ers are regul arl y i nvi ted to parti ci pate i n the audi -
ti oni ng and sel ecti on of cabi n attendants. And
many take ti me off from work to joi n Southwests
empl oyee sel ecti on team as i t carri es out i ts work.
As one customer commented, Why not do i t? Its
my ai rl i ne.
18. To what extent are measures of customer satis-
faction, customer loyalty, or the quality and quan-
tity of service output used in recognizing and re-
warding employees?
Empl oyee recogni ti on may often i nvol ve l i ttl e
more than i nformi ng i ndi vi dual empl oyees or em-
pl oyees as a group about servi ce i mprovements and
i ndi vi dual successes. Banc One goes one step fur-
ther, i ncl udi ng customer sati sfacti on measures for
each banki ng uni t i n i ts peri odi c report of other per-
formance measures, mostl y fi nanci al , to al l uni ts.
Internal Service Quality
19. Do employees know who their customers are?
It i s parti cul arl y di ffi cul t for empl oyees to i denti -
fy thei r customers when those customers are i nter-
nal to the company. These empl oyees often do not
THE SERVICE-PROFIT CHAIN
174 HARVARD BUSINESS REVIEW March-Apri l 1994
know what i mpact thei r work has on other depart-
ments. I denti fyi ng i nter nal customers requi res
mappi ng and communi cati ng characteri sti cs of
work fl ow, organi zi ng peri odi c cross-departmental
meeti ngs between customers and servers, and
recogni zi ng good i nternal servi ce performance.
In 1990, USAA organi zed a PRIDE (Professi onal -
i sm Resul ts i n Dedi cati on to Excel l ence) team of
100 empl oyees and managers to exami ne and i m-
prove on a functi on-by-functi on basi s al l processes
associ ated wi th property-and-casual ty i nsurance
admi ni strati on, whi ch i ncl uded anal yzi ng cus-
tomer needs and expectati ons. The PRIDE effort
was so successful that i t l ed to a cross-functi onal re-
vi ew of USAAs servi ce processi ng. Servi ce process-
i ng ti me has been reduced, as have handoffs of cus-
tomers from one server to another.
20. Are employees satisfied with the technological
and personal support they receive on the job?
The cornerstone of success at Taco Bel l i s the pro-
vi si on of the l atest i n i nformati on technol ogy, food
servi ce equi pment, si mpl e work-schedul i ng tech-
ni ques, and effecti ve team trai ni ng. Thi s practi ce
l ed to the establ i shment of sel f-managi ng teams of
servi ce provi ders. Al so, the qual i ty of work l i fe i n-
vol ves sel ecti ng the ri ght workers. Wi nners l i ke to
be associ ated wi th wi nners. Better empl oyees tend
to refer peopl e they l i ke and peopl e l i ke themsel ves.
Internal servi ce qual i ty can al so be thought of as
the qual i ty of work l i fe. It i s a vi si bl e expressi on of
an organi zati ons cul ture, one i nfl uenced i n i mpor-
tant ways by l eadershi p.
Leadership
21. To what extent is the companys leadership:
a. energetic, creative vs. stately, conservative?
b. participatory, caring vs. removed, elitist?
c. listening, coaching, and teaching vs. supervis-
ing and managing?
d. motivating by mission vs. motivating by fear?
e. leading by means of personally demonstrated
values vs. institutionalized policies?
22. How much time is spent by the organizations
leadership personally developing and maintaining
a corporate culture centered around service to cus-
tomers and fellow employees?
Leaders natural l y hav e i ndi v i dual trai ts and
styl es. But the CEOs of compani es that are success-
ful l y usi ng the servi ce-profi t chai n possess al l or
most of a set of trai ts that separate them from thei r
merel y good competi tors. Of course, di fferent
styl es of l eadershi p are appropri ate for vari ous
stages i n an organi zati ons devel opment. But the
messages sent by the successful l eaders we have
observed stress the i mportance of careful attenti on
to the needs of customers and empl oyees. These
l eaders create a cul ture capabl e of adapti ng to the
needs of both.
Relating the Measures
23. What are the most important relationships in
your companys service-profit chain?
24. To what extent does each measure correlate
with profit and growth at the frontline level?
25. I s the importance of these relationships reflect-
ed in rewards and incentives offered to employees?
Measures dri ve acti on when they are rel ated i n
ways that provi de managers wi th di recti on. To en-
joy the ki nd of success that servi ce organi zati ons
l i ke Southwest Ai rl i nes, Servi ceMaster, and Taco
Bel l have enjoyed, l ooki ng at i ndi vi dual measures i s
not enough. Onl y i f the i ndi vi dual measures are
ti ed together i nto a comprehensi ve pi cture wi l l the
servi ce-profi t chai n provi de a foundati on for un-
precedented profi t and growth.
Repri nt 94204