,Matching Supp n
ly with Demand
n n
An Introduction to Oper
n n n
ations Management
n
Fourth Edition
n
Gérard Cachon
n
ThenWhartonnSchool,nUniversit
ynofnPennsylvania
Christian Terwiesch
n
ThenWhartonnSchool,nUniversit
ynofnPennsylvania
,Preface
Thisnbo ok nrep res ent s nourn vie w n ofnthenes sen tial nb ody nofnkn owl edg e nfor nannint rod uct ory nop er-
n ation s n mana gem ent n c our se. n I t n has n be en n su cce ssf ully n u sed n w ith n a ll n ty pes n of n s tud ent s, n fr om n fr es hm e
nnta ki ng n a nnin tr od uc to ry n co ur se ni nnop er at io ns n ma na gem en t, n tonM BA s, nt one xe cu ti ve nMBAs,nandnev
en nPhD nstudents.
Our nguiding nprinciple nin nthe ndevelopment nof nMatch ing n Supply n with n De mand n has nbeen n“rea
lnoperations,nreal nsolutions.”n“Realnoperations”nmeans nthatnmostnofnthenchaptersninnthisnbookn arenwri tte n
n from n th e n pers pec tiv e n of n a n sp ecif ic n co mpa ny n so n t hat n th e n mat eri al n in n t his n tex t n will n come n to n life n by n di
scussingn itninn anreal-
worldnconte xt. nCompan ies n andnprod ucts nare ns i mp ly n e as ie r nt on re me m be r nt h an n nu m be rs n a ndn eq ua t io n s.
n We n ha ve n ch o se n n a n wi d e n v ar i et y n of n c omp ani es, n sm all n a nd n la rge , n rep rese nti ng n se rvi ces , n manu fac tur
ing,na ndn ret ail ing na lik e. nW hi le no bv io us ly nn ot nf ul ly n re pr es en ta ti ve , nwe nb el ie ve n th at —
tak en n to ge th er —
thesencases np ro vi d e nanr e al i st ic np ic t ur e nof no p er at i on s n ma n ag e me nt np r ob le m s nto da y .
“R e al ns ol u ti on s ”nme a ns nt ha t nwend onn o t nw a nt ne qu a ti o ns na nd nm od el s nto nm e re ly np ro v id e nst ude nts n
withnm athe mat ica l ngym nas tic s n fornth e nsak e nofn annin tel lec tua l nexe rci se. nWenf eel nthat nprofessional nt
raining,n evennin nanrigorous nacademicn setting,nrequiresntools nandnstrategies nthatns tu de n ts n c an ni mp l em
en tn innp ra c ti ce . nWe na c hi ev e nth i s nb ynde m on s tr a ti ng nh ow nt onap pl y no ur nmod els nfr om nsta rt ntonfin ish ni
nnanrealis tic nop era tio nal nse tti ng. n Fu rth erm ore , nwe nopen ly nad dre ss ntheni mpl eme nta tion nc hal len ges n
ofneac h nmode l/s tra teg y nwendi scus s nsonth at nst ude nts nknow nw hat nt o ne xpect nw hen nt he n“rubber nh its nt he n
pavement.”
To nfullyndeliver nonn“realnoperations,nreal nsolutions,”nwenalsonmust nadherento nthenprin-
n ciple nof n“realnsimple.”nDonnot n worry;n “real n simple”n does n notn mean n plentyn of n “blah-
blah”nwithoutnany nanalyticaln rigor.n Quite nthencontrary. nTonus,n“realnsimple”nmeansn hardn analysisnthatn i
snmadeneasyntonlearn.nThisnisncrucial nfornannoperations ntext.nOurnobjectivenis ntonteachnbusine ss nle ade rs
,nnotnta ctic ian s. nTh us, nw enneed ns tud ent s ntonbe na ble ntonq uic kly nd eve lop nanfou ndatio n nofnformaln mode l
snsonthatn theynhave nthe ntimenton explo re nthe nbignpictu re, nthatnis, nh own o pe ra ti on s nc ann b entr an sf or me d n
tonp ro vi de na nn or ga ni za ti on n wi th ns us ta in ab le nc om pe ti -
n tive n ad vant age n a nd/ or n su per ior n cu sto mer n s erv ice . n Stud ent s n who n g et n bo gged n d own n i n n det ail s, n equati
ons,nandnanalysis narennotnfullyncapturingnthenvaluable ninsightsntheynwillnneed ninntheir nfuture ncareer
.
So nhow ndonwe nstrive nforn“real nsimple”?nFirst, nwenrecognize nthatnnot neverynstudentncomes n tonthisnmate
rialnwithnannengineering/math nbackground. nAsnan result,nwentriedn tonusenasnlittle n mathemat ical nn ota tio
nnasnpos sibl e, ntonp rov ide nm any nre al-
worl d nexam ple s, nan d ntonad her e n toncon sis ten t nter mino log y nandnp hra sin g. nSec ond , nwen pr ovi de nvar iou s n
leve ls nof nde tai l nfor ne ach nanalysis. nFornexample,neverynlittle nstepninnannanalysisnisndescribedninnthentext
nvianannexplicitn example; n thenn a n summaryn of n then process n isn provided n inn a n “hown to”n exhibit,n a n briefn listing
n of n keynnotation nandnequations nisn provided n atn the n endn of n each n chapter,n and, n finally,n solved n prac-
n tice n pro ble ms n ar e n offe red n t o n rei nfo rce n le arn ing . n Whi le n we n do n h umb ly n re cog niz e, n giv en n th e n quantita
tivensophistication nofnthisn text,n thatn “muchn simpler”n mightn ben moren accuratenthann“realn simple,”n wen n
evertheless n hopen that nstudentsnwilln benpleasantly nsurprisedntondiscovernthat nthei r nanaly tic al ncap abi
litiesnareneven nstronger nthannth ey nima gin ed.
Th e nin i ti al nv er s io n nof nMatching n Supplyn withn Demand n mad e nit s nde bu t nin np or t io n s no fnth en o pera
tion s nmana gem ent nc ore nc our se nat nWh art on n innt he n200 2–2003 na cad emi c nyea r. nThi s nedi -
n tion n inc orp ora tes n th e n feed bac k n we n have n re cei ved n ov er n the n l ast n 16 n ye ars n fr om n man y n stud ent s, n execut
ives, na nd nc olleagues, nb oth na t nW harton na nd na broad.
Gérard n Cachon
Christian n Terwiesch
ix
, Changes to This Edition n n n
The nfourth nedition nhas nbenefited nfrom nthe ncomments nand nsuggestions nfrom nstudents, nf a cu lt
y,nandnprac ti ti o ne r s nfr o m na ro un d nt h e nwo r ld .
Thenimple me n te d nch an g es nc an nb e nd i vi de d nin t o nt h re e nca t eg or i es : nannu pd at e nof nd at a nan d nc a se ne x
amples,nt he na dd i ti o n no fn tw onch a pt e rs nr el at e d ntonc on te n t nt h at nw as nn ot np re v io u sl y nco ver ed ni nnth en
book, nandnannoveral l nst ream lin ing nof nt hene xpo sit ion n ofnt he ne xis tin g nco nten t. nThe nworld nhas nchange
d nagain nbetween nthis nand nthe nprevious nedition. nRide nsharing, napart men t nsha rin g, nan d nele ctr i
cnvehicles nwere nnotnyet nant hin g n“ba ck nth en. ”nCons equen tly,
wenhave nup da t ed nda t a na n d ne x am pl e s nt ontr y nt onma i nt ai n nth e nti m el in e ss no fnt he nc on te nt.
We nhav e nadde d ntwon n ewnch apt ers nton t his n b ook . nThenfi rst n n ewnch apt er nisnab out n f ore cas t-
n ing, n whi ch n is n an n ab solu tel y n esse nti al n inp ut n to n al l n oper ati ons n mo del s. n The n gr owt h n of n avai l-
n able n da ta n onl y n mak es n fo rec asti ng n mo re n re lev ant . n The n se con d n new n c hap ter n is n o n n sch edu lin g. n We n co ver
ednsch edul ing ni n near ly nte xts , nbutnno t ntonth e next ent nth e ntop ic nde ser ves ngi ven no ur nc o nt in u ed n em p ha s
isn onn se r vi c e no pe ra t io n s. n No w nw enp r ov i de n and ed i ca t ed n an d nm o re n ex te n -
n sive ncoverage nof nscheduling.
We nhav e nmaden ann umb er nof ns mal l nchan ges n t hat nm ake nth e nmat eri al nea sie r nfornst ude nts nt onabs orb. nF o
rnexam ple, nw enhav e nstr eam lin ed n thene xpo sit ion no f nlabo r nuti liz ati on n cal cul ati ons n an d nw enha v e nde -
em p ha si z ed nt h e nus e no fnth e ne xp e ct e d n lo s s n fu n ct io n ni n nt he nn e ws ve n do r na nd norder-up-
tonmodels.n Insteadnofn thenlossnfunction, nwen providen the n “expectedn inventoryn func-
n tion,”n w hich n al low s n stud ent s n to n arri ve n at n the n ne ces sar y n answe r n with n fe wer n st eps . n Furt her -
n more, n we n fi nd n tha t n stud ent s n are n ab le n to n int uit ive ly n gra sp n wha t n the n inv ent ory n fu nct ion n do es n b e tt er n
thannthenlo s s nf un ct i on .
x
ly with Demand
n n
An Introduction to Oper
n n n
ations Management
n
Fourth Edition
n
Gérard Cachon
n
ThenWhartonnSchool,nUniversit
ynofnPennsylvania
Christian Terwiesch
n
ThenWhartonnSchool,nUniversit
ynofnPennsylvania
,Preface
Thisnbo ok nrep res ent s nourn vie w n ofnthenes sen tial nb ody nofnkn owl edg e nfor nannint rod uct ory nop er-
n ation s n mana gem ent n c our se. n I t n has n be en n su cce ssf ully n u sed n w ith n a ll n ty pes n of n s tud ent s, n fr om n fr es hm e
nnta ki ng n a nnin tr od uc to ry n co ur se ni nnop er at io ns n ma na gem en t, n tonM BA s, nt one xe cu ti ve nMBAs,nandnev
en nPhD nstudents.
Our nguiding nprinciple nin nthe ndevelopment nof nMatch ing n Supply n with n De mand n has nbeen n“rea
lnoperations,nreal nsolutions.”n“Realnoperations”nmeans nthatnmostnofnthenchaptersninnthisnbookn arenwri tte n
n from n th e n pers pec tiv e n of n a n sp ecif ic n co mpa ny n so n t hat n th e n mat eri al n in n t his n tex t n will n come n to n life n by n di
scussingn itninn anreal-
worldnconte xt. nCompan ies n andnprod ucts nare ns i mp ly n e as ie r nt on re me m be r nt h an n nu m be rs n a ndn eq ua t io n s.
n We n ha ve n ch o se n n a n wi d e n v ar i et y n of n c omp ani es, n sm all n a nd n la rge , n rep rese nti ng n se rvi ces , n manu fac tur
ing,na ndn ret ail ing na lik e. nW hi le no bv io us ly nn ot nf ul ly n re pr es en ta ti ve , nwe nb el ie ve n th at —
tak en n to ge th er —
thesencases np ro vi d e nanr e al i st ic np ic t ur e nof no p er at i on s n ma n ag e me nt np r ob le m s nto da y .
“R e al ns ol u ti on s ”nme a ns nt ha t nwend onn o t nw a nt ne qu a ti o ns na nd nm od el s nto nm e re ly np ro v id e nst ude nts n
withnm athe mat ica l ngym nas tic s n fornth e nsak e nofn annin tel lec tua l nexe rci se. nWenf eel nthat nprofessional nt
raining,n evennin nanrigorous nacademicn setting,nrequiresntools nandnstrategies nthatns tu de n ts n c an ni mp l em
en tn innp ra c ti ce . nWe na c hi ev e nth i s nb ynde m on s tr a ti ng nh ow nt onap pl y no ur nmod els nfr om nsta rt ntonfin ish ni
nnanrealis tic nop era tio nal nse tti ng. n Fu rth erm ore , nwe nopen ly nad dre ss ntheni mpl eme nta tion nc hal len ges n
ofneac h nmode l/s tra teg y nwendi scus s nsonth at nst ude nts nknow nw hat nt o ne xpect nw hen nt he n“rubber nh its nt he n
pavement.”
To nfullyndeliver nonn“realnoperations,nreal nsolutions,”nwenalsonmust nadherento nthenprin-
n ciple nof n“realnsimple.”nDonnot n worry;n “real n simple”n does n notn mean n plentyn of n “blah-
blah”nwithoutnany nanalyticaln rigor.n Quite nthencontrary. nTonus,n“realnsimple”nmeansn hardn analysisnthatn i
snmadeneasyntonlearn.nThisnisncrucial nfornannoperations ntext.nOurnobjectivenis ntonteachnbusine ss nle ade rs
,nnotnta ctic ian s. nTh us, nw enneed ns tud ent s ntonbe na ble ntonq uic kly nd eve lop nanfou ndatio n nofnformaln mode l
snsonthatn theynhave nthe ntimenton explo re nthe nbignpictu re, nthatnis, nh own o pe ra ti on s nc ann b entr an sf or me d n
tonp ro vi de na nn or ga ni za ti on n wi th ns us ta in ab le nc om pe ti -
n tive n ad vant age n a nd/ or n su per ior n cu sto mer n s erv ice . n Stud ent s n who n g et n bo gged n d own n i n n det ail s, n equati
ons,nandnanalysis narennotnfullyncapturingnthenvaluable ninsightsntheynwillnneed ninntheir nfuture ncareer
.
So nhow ndonwe nstrive nforn“real nsimple”?nFirst, nwenrecognize nthatnnot neverynstudentncomes n tonthisnmate
rialnwithnannengineering/math nbackground. nAsnan result,nwentriedn tonusenasnlittle n mathemat ical nn ota tio
nnasnpos sibl e, ntonp rov ide nm any nre al-
worl d nexam ple s, nan d ntonad her e n toncon sis ten t nter mino log y nandnp hra sin g. nSec ond , nwen pr ovi de nvar iou s n
leve ls nof nde tai l nfor ne ach nanalysis. nFornexample,neverynlittle nstepninnannanalysisnisndescribedninnthentext
nvianannexplicitn example; n thenn a n summaryn of n then process n isn provided n inn a n “hown to”n exhibit,n a n briefn listing
n of n keynnotation nandnequations nisn provided n atn the n endn of n each n chapter,n and, n finally,n solved n prac-
n tice n pro ble ms n ar e n offe red n t o n rei nfo rce n le arn ing . n Whi le n we n do n h umb ly n re cog niz e, n giv en n th e n quantita
tivensophistication nofnthisn text,n thatn “muchn simpler”n mightn ben moren accuratenthann“realn simple,”n wen n
evertheless n hopen that nstudentsnwilln benpleasantly nsurprisedntondiscovernthat nthei r nanaly tic al ncap abi
litiesnareneven nstronger nthannth ey nima gin ed.
Th e nin i ti al nv er s io n nof nMatching n Supplyn withn Demand n mad e nit s nde bu t nin np or t io n s no fnth en o pera
tion s nmana gem ent nc ore nc our se nat nWh art on n innt he n200 2–2003 na cad emi c nyea r. nThi s nedi -
n tion n inc orp ora tes n th e n feed bac k n we n have n re cei ved n ov er n the n l ast n 16 n ye ars n fr om n man y n stud ent s, n execut
ives, na nd nc olleagues, nb oth na t nW harton na nd na broad.
Gérard n Cachon
Christian n Terwiesch
ix
, Changes to This Edition n n n
The nfourth nedition nhas nbenefited nfrom nthe ncomments nand nsuggestions nfrom nstudents, nf a cu lt
y,nandnprac ti ti o ne r s nfr o m na ro un d nt h e nwo r ld .
Thenimple me n te d nch an g es nc an nb e nd i vi de d nin t o nt h re e nca t eg or i es : nannu pd at e nof nd at a nan d nc a se ne x
amples,nt he na dd i ti o n no fn tw onch a pt e rs nr el at e d ntonc on te n t nt h at nw as nn ot np re v io u sl y nco ver ed ni nnth en
book, nandnannoveral l nst ream lin ing nof nt hene xpo sit ion n ofnt he ne xis tin g nco nten t. nThe nworld nhas nchange
d nagain nbetween nthis nand nthe nprevious nedition. nRide nsharing, napart men t nsha rin g, nan d nele ctr i
cnvehicles nwere nnotnyet nant hin g n“ba ck nth en. ”nCons equen tly,
wenhave nup da t ed nda t a na n d ne x am pl e s nt ontr y nt onma i nt ai n nth e nti m el in e ss no fnt he nc on te nt.
We nhav e nadde d ntwon n ewnch apt ers nton t his n b ook . nThenfi rst n n ewnch apt er nisnab out n f ore cas t-
n ing, n whi ch n is n an n ab solu tel y n esse nti al n inp ut n to n al l n oper ati ons n mo del s. n The n gr owt h n of n avai l-
n able n da ta n onl y n mak es n fo rec asti ng n mo re n re lev ant . n The n se con d n new n c hap ter n is n o n n sch edu lin g. n We n co ver
ednsch edul ing ni n near ly nte xts , nbutnno t ntonth e next ent nth e ntop ic nde ser ves ngi ven no ur nc o nt in u ed n em p ha s
isn onn se r vi c e no pe ra t io n s. n No w nw enp r ov i de n and ed i ca t ed n an d nm o re n ex te n -
n sive ncoverage nof nscheduling.
We nhav e nmaden ann umb er nof ns mal l nchan ges n t hat nm ake nth e nmat eri al nea sie r nfornst ude nts nt onabs orb. nF o
rnexam ple, nw enhav e nstr eam lin ed n thene xpo sit ion no f nlabo r nuti liz ati on n cal cul ati ons n an d nw enha v e nde -
em p ha si z ed nt h e nus e no fnth e ne xp e ct e d n lo s s n fu n ct io n ni n nt he nn e ws ve n do r na nd norder-up-
tonmodels.n Insteadnofn thenlossnfunction, nwen providen the n “expectedn inventoryn func-
n tion,”n w hich n al low s n stud ent s n to n arri ve n at n the n ne ces sar y n answe r n with n fe wer n st eps . n Furt her -
n more, n we n fi nd n tha t n stud ent s n are n ab le n to n int uit ive ly n gra sp n wha t n the n inv ent ory n fu nct ion n do es n b e tt er n
thannthenlo s s nf un ct i on .
x
