HowFinancialFirmsDecideonTechnology，介紹國際大銀行在決定對信息技術投資時的考慮要點和他們具體的實施過程。HowFinancialFirmsDecideonTechnology(Abstract)Thefinancialservicesindustryisthemajorinvestorininformationtechnology(IT)intheU.S.economy;thetypicalbankspendsasmuchas15%ofnon-interesteexpensesonIT.Apersistentfindingofresearchintotheperformanceoffinancialinstitutionsisthatperformanceandefficiencyvarywidelyacrossinstitutions.NowhereisthisvariabilitymorevisiblethanintheoutcomesoftheITinvestmentdecisionsintheseinstitutions.ThispaperpresentstheresultsofanempiricalinvestigationofITinvestmentdecisionprocessesinthebankingindustry.Thepurposeofthisinvestigationistouncoverwhat,ifanything,canbelearnedfromtheITinvestmentpracticesofbanksthatwouldhelpinunderstandingthecauseofthisvariabilityinperformancealongwithpointingtowardmanagementpracticesthatleadtobetterinvestmentdecisions.UsingPCbankingandthedevelopmentofcorporateInternetsitesasthecasestudiesforthisinvestigation,thepaperreportsondetailedfield-basedsurveysofinvestmentpracticesinseveralleadinginstitutionsHowFinancialFirmsDecideonTechnology(PartOne)1.0IntroductionInformationtechnology(IT)isincreasinglycriticaltotheoperationsoffinancialservicesfirms.Todaybanksspendasmuchas15%ofnon-interestexpenseoninformationtechnology.Itisestimatedthattheindustrywillspendatleast$21.1billiononITin1998,andfinancialinstitutionscollectivelyaccountforthemajorityofITinvestmentintheU.S.economy.Inadditontobeingalargecomponentofthecoststructure,informationtechnologyhasastronginfluenceonfinancialfirmsoperatonsandstrategy.Fewfinancialproductsandservicesexistthatdonotutilizecomputersatsomepointinthedeliveryprocess,andafirms'informationsystemsplacestrongconstraintsonthetypeofproductsoffered,thedegreeofcustomizationpossibleandthespeedatwhichfirmscanrespondtocompetitiveopportunitiesorthreats.Apersistentfindingofresearchintotheperformanceoffinancialinstitutionsisthatperformanceandefficiencyvarieswidelyacrossinstitutions,evenaftercontrollingforfactorssuchassize(scale),productbreadth(scope),branchingbehaviorandorganizationalform(e.g.stockversusmutualforinsurers;banksversussaving&loans).Giventhecentralrolethattechnologyplaysintheseinstitutions,atleastsomeofthisvariationislikelytobeduetovariationsintheuseandeffectivenessofITinvestments.WhilesomeauthorshavearguedthatthevalueofITinvestmenthasbeeninsignificant,particularlyinservices,recentempiricalworkhassuggestedthatITinvestment,onaverage,isaproductiveinvestment.Perhapsmoreimportantly,thereappearstobesubstantialvariationacrossfirms;somefirmshaveveryhighinvestmentsbutarepoorperformers,whileotheresinvestlessbutappeartobemuchmoresuccessful.BrynjolfssonandHittfoundthatasmuchashalfthereturnstoITinvestmentareduetofirmspecificfactors.OnepotentiallyimportantdriverofdifferencesinITvalue,andoffirmperformancemorebroadly,islikelytobethedecisionandmanagementpeocessedforITinvestments.HorrorstoriesofbadITinvestmentdecisionsabound.Considertheexampleofthenewstrategicbankingsystem(SBS)atBancOne(AmericanBanker1997).BancOneCorp.andElectronicDataSystemsCorp.agreedlastyeartoendtheirjointdevelopmentofthisretailbankingsystemafterspendinganestimated$175milliononit.AsstatedintheAmericanBankerarticle,SBS"wasjustsooverwhelmingandsocompletethatbythetimetheyweregettingtomarket,itwasgoingtotaketoolongtoinstallthewholething,"saidAlanRiegler,principalinErnst&Young'sfinancialservicesmanagementconsultingdivision.However,notallthestoriesarenegative.NewITsystemsareplayingavitalroleinreshapingthedeliveryoffinancialservices.Forexample,newcomputer-telephonyintegration(CTI)technologiesaretransformingcallcenteroperationsinfinancialinstitutions.Byinvestingintechnology,moreandmoreinstitutionsaremovingoperationsfromhigh-costbranchoperationstothetelephonechannel,wherethecostpertransactionisone-tenththecostofatellerinteraction.ThisITinvestmentnotonlyreducesthecostofservingexistingcustomers,butalsoextendsthereachoftheinstitutionbeyonditstraditionalgeographicboundaries.Inthispaper,weutilizedetailedcasestudiesofsixretailbankstoinvestigateseveralinterrelatedquestions:１.WhatprocessesdobanksutilizetoevaluateandmanageITinvestments?

２.HowwelldoactualpracticesalignwiththeoreticalargumentsabouthowITinvestmentsshould

bemanaged?

３.WhatimpactdoesthatmanagementofITinvestmentshaveonperformance?HowFinancialFirmsDecideonTechnology(PartTwo)Forthefirstquestion,wedevelopastructuredframeworkforcatalogingITinvestmentpracticesandthenpopulatethisframeworkusingacombinationofsurveysandsemi-structuredinterviews.WethencomparetheresultsofthisexercisewithasynthesisoftheliteratureonITdecisionmakingtounderstandinghowpracticesvaryacrossfirmsandtheextenttowhichthisisconsistentwith"bestpractices"asdescribedinpreviousliterature.Finally,wewillcomparetheseprocessestointernalandexternalperformancemetricstobetterunderstandwhichsetsofpracticesappeartobemosteffective.Tomakethesecomparisonsconcrete,weexamineboththegeneraldecisionprocessaswellasthespecificprocessesusedfortworecentITinvestmentdecisions:theadoptionofcomputer-basedhomebanking(PCbanking),andthedevelopmentofthecorporatewebsite.Thesedecisionswerechosenbecausetheywererecentandarerelatedbutprovidesomecontrast;inparticular,PCbankingisafairlywelldefinedproductinnovation,whilethecorporatewebpresenceismoreofaninfrastructureinvestmentwhichislesswell-definedintermsofobjectivesandbusinessownership.Overall,wefindthatwhilesomeaspectsofthedecisionprocessarefairlysimilaracrossinstitutionsandoftenconformto"bestpractice"asdefinedbypreviousliterature,thereareseveralareaswherethereislargevariationinpracticeamongthebanksandbetweenactualandtheoreticalbestpractice.Mostbankshaveastrongandstandardizedprojectmanagementforongoingsystemsprojects,andformalstructuresforinsuringthatline-managersandsystemspeopleareincontactattheinitiationoftechnologyprojects.Atthesametime,manybankshaverelativelyweakprocesses(bothformalandinformal)foridentifyingnewITinvestmentopportunities,allocatingresourcesacrossorganizationallines,andfundingexploratoryorinfrastructureprojectswithlongtermoruncertainpayoffs.Thereminderofthispaperisorganizedasfollows.Section2describesthepreviousliteratureonperformanceoffinancialinstitutionsandtheeffectsofITonperformance.Section3describesthemethodsanddata.Section4describesthecurrentacademicthinkingonvariouscomponentsofthedecisionprocessandcomparesthattoactualpracticesatthebankswevisited.Section5describestheresultsofourin-depthstudyofPCbankingprojectsandthesummary,Section6containsasimilaranalysisfortheCorporateWebSiteanddiscussionandconclusionappearinSection7.三Ho數(shù)w義Fi特na母nc燃ia恢l隔Fi冊rm雄s法De革ci蓄de火o卸n里Te否ch撥no鈴lo仰gy延(削Pa謎rt慣T螞hr美ee壯)肉2.謀0蓋P仇re肝vi毫ou稿s隱L認it谷er矛at言ur豎e

第2.吵1耗P勻er豈fo坑rm翅an翼ce蹄o栗f柱Fi旋na享nc舌ia跡l以In退st況it棋ut莖io昏ns

釘克Th漂er譽e當ha抖ve差b項ee轎n影a愛nu浸mb勁er縫o邊f(xié)在st好ud卻ie黃s尊th淘at泄h戴av繡e撒ex怕am暴in罰ed陰t浙he扭e漸ff遮ic葛ie般nc帖y桶of吧t依he摧b該an竟ki舍ng捆達in掀du短st記ry下a體nd好泉th迷e希ro會le置o逐f衰va遷ri姿ou百s縮fa沖ct贈or煉s斑su臥ch對a女s幼co忽rp鄭or臉at折e犯co摘nt汪ro恐l奉st主ru辣ct禾ur參e憑(拿ty寄pe脈o條f廉bo旁ar咽d,禾d消ir頭ec樸to度rs遞,袍in回si巷de棚r添st階oc駐k餓ho就ld告in內gs覺,勸et聞c.布),斜e鋒co艦no你mi研es裝o剖f理sc抄al場e墓(s貢iz眉e)藝,芳ec澡on帳om疾ie器s銳of浙s午co纏pe綁(叔pr腿od戲uc鄰t駛br盒ea刻dt宿h)墓,互an召d鑼br另an命ch傳in艙g星st喘ra鋼te械gy張;墾se拉e渡Be匆rg接er頸,寨Ka答sh命yu討p憤an著d唯Sc旅al倒is茶e母(1罵99堂5)街a價nd斑H路ar蓄ke攔r調an亡d耽Ze裂ni安os隨(歇fo緊rt岔hc蹲om獨in涂g)腫f摔or限a均r慕ev徒ie靜w餐of弄t謀he享b庸an學ki扇ng絨e巖ff究ic捏ie穿nc滋y缸li咬te鴨ra池tu延re獵.上Wh己il程e數(shù)th耐er毛e旺is碰s椅ub片st未an遼ti癢al雞d銀eb到at短e含as好t望o疫th異e擱ro培l(xiāng)e科o康f瓦th賴es親e眾va題ri選ou蘋s煩fa裹ct臭or拘s,己t無he較re啞i竭s獻on范e月un惑am忍bi換gu輔ou久s浩re姻su繼lt牲:往th熔at病m匠os截t鑰of餓t痛he喘(閱in侄)召ef滿fi片ci矮en憐cy呼o研f臘ba嗓nk勞s滿is百n沸ot循e泄xp志la袍in菊ed抵b筒y撥th峰e克fa勒ct筋or見s掀th虹at疏h冒av匙e濕be瓣en淹c雹on諷si爛de吳re變d莖in皆p鍵ri判or偏w去or寄k.娛F儲or昏e林xa必mp灑le衰,離Be母rg壁er白a鹽nd共M鋒es械te鮮r啦(1仿99櫻7)顏e毀st另im離at晶e柔t(yī)h斯at聲a囑s南mu墊ch速a頂s伯65蜜-9屬0%欣o泊f險th癥e流x-蛛in鑼ef羨fi勉ci贏en禾cy飛r傾em究ai蠟ns擇u睡ne斤xp抬la械in瘋ed陽a踐ft催er液c池on春tr帝ol撐li秒ng甲f悅or早k降no霞wn填d中ri貝ve顫rs社o荷f峰pe慌rf轎or墊ma音nc波e.京A籃s飾im申il單ar磁s另to龜ry敗a伶ls哥o陶ap遞pe樂ar研s經in轟i桐ns洗ur惱an友ce控w濾he停re震"擠x-串ef朝fi暫ci駱en隊cy星"詢va憤ri澆es戴s命ub寒st諸an諷ti推al枯ly陶a寄cr逮os贏s撓fi語rm詠s廟wh刻en涂s需iz察e,吵s籮co稿pe握,裳pr扒od奴uc揭t脖mi毛x,義d碼is容tr白ib資ut臘io尊n淡st辰ra胡te品gy況a秧nd蒙o竿th歇er急s岸tr屋at度eg鴨ic石v傍ar號ia貌bl旨es銹a擾re根c滾on兼si偽de廊re沃d.秧I吧t榆ha瞞s障be撇en摔a露rg臉ue著d元th莊at務o提ne瞧m勝us懂t頸ge腐t歪"i拉ns愛id較e類th今e誘bl廊ac漸k淡bo膝x"尋o待f標th銀e胸ba響nk采o響t考co咳ns凝id舞er輔t怕he鉛r器ol停e宮of踏o茄rg碰an窩iz課at字io于na助l,痕s噸tr玻at歷eg享ic醬a鍛nd姓t朱ec逢hn肉ol御og縱ic凳al窗f滋ac暈to暈rs澇t攔ha督t抬ma擱y哈be換m機is癥se劇d飼in轟s灰tu鐵di活es慢t陜ha聲t擔re船ly叼h腦ea吧vi捎ly添o命n查pu診bl剝ic轎f萬in伯an興ci黑al景d派at闖a.2.2InformationTechnologyandBusinessValue

EarlystudiesoftherelationshipbetweenITandproductivityorothermeasuresofperformanceweregenerallyunabletodeterminethevalueofITconclusively.Loveman(1994)andStrassmann(1990),usingdifferentdataandanalyticalmethodsbothfoundthattheperformanceeffectsofcomputerswerenotstatisticallysignificant.Barus,KriebelandMukadopadhyay(1995),usingthesamedataasLoveman,foundevidencethatITimprovedsomeinternalperformancemetricssuchasinventorytrunover,butcouldnottiethesebenefitstoimprovementsinbottomlineproductivity.Althoughthesestudieshadanumberofdisadvantages(smallsamples,noisydata)whichyieldedimprecisemeasuresofITeffects,thislackofevidencecombinedwithequallyequivocalmacroeconomicananlysesbyStevenRoach(1987)implicitlyformedthebasisforthe"productivityparadox".AsRobertSolow(1987)onceremarked,"youcanseetehcomputerageeverywhereexceptintheproductivitystatistics."

MorerecentworkhasfoundthatITinvestmentisasubstantialcontributortofirmproductivity,productivitygrowthandstockmarketvaluationinasamplethatcontainsawiderangeofindustries.BrynjolfssonandHitt(1994,1996)andLichtenberg(1995)foundthatITinvestmenthadapositiveandstatisticallysignificantcontributiontofirmoutput.BrynjolfssonandYang(1997)foundthatthemarketvaluationofITcapitalwasseveraltimesthatofordinarycapital.BrynjolfssonandHittalsofoundastrongrelationshipbetweenITandproductivitygrowthandtahtthisrelationshipgrowsstrongeraslongertimeperiodsareconsidered.Collectively,thesestudiessuggestthatthereisnoproductivityparadox,atleastwhentheanalysisisperformedacrossindustriesusingfirm-leveldata.Thedifferencesbetweentheseresultsandearlierstudiesisprobablyduetotheuseofdatatahtwasrecent,morecomprehensice,andmoredisaggregated(firmlevelratherthanindustryoreconomylevel).

Mostprevioussutdieshaveconsideredtheeffectsoftechnologyacross刃f泰ir冊ms啦i毛n門mu枕lt和ip映le庸i那nd逃us肺tr英ie皺s,共a岔lt攪ho涼ug膊h醒a濾fe眠w鑄st拾ud舒ie妙s銷ha繪ve說c伙on眉si鄭de勢re貸d絞th寧e伯ro挖le償o頂f頃te甩ch膜no浙lo湊gy言i茅n星sp小ec子if麥ic筆al集ly駁i將n朽th觀e毅ba罰nk毫in宴g敏in載du朗st強ry還.容St袖ei必ne在r汪an糕d秀Te深ix間ie寶ra脫s山ur波ve勁ye毛d輕th床e母ba痛nk郊in效g霞in伙du展st勵ry祝a姐nd默a世rg腦ue獨d漲th悔at光w株hi使le譜l灶ar而ge岡i亡nv查es訴tm球en蛙ts挎i饒n摔te蛇ch貓no斬lo甲gy拖c律le爸ar噸ly毛h丹ad誠v該al子ue肝,l約it莫tl用e首of揮t鍛hi抗s深va收lu拿e陰wa吸s滾be天in杏g戲ca得pt流ur顏ed拾b呀y(tǒng)卵th技e圓ba關nk寫s甘th鑄em動se劃lv摧es范;并mo暮st雖o炮f錫th帝e印be品ne昂fi題ts丈w結er培e須be造in草g烏pa休ss安ed抬o寄n婦to燈c教us永to默me勿rs竊a法s恭a虛re觀su江lt稠o涂f怨in顯te判ns咐e誓co戴mp偷et閱it練io墨n.拍A晉lp李ar煙a冒nd乳K艇im棗e外xa折mi慌ne姐d霉th篩e旱co紹st目e婦ff土ic肺ie僻nc錫y希of率b伶an舞ks奏o凝ve濁ra限ll慰a手nd秧f疼ou蜘nd赤t咬ha息t圓IT惡i失nv行es功tm則en飽t握wa放s偷as喪so挑ci沸at勒ie排d反wi監(jiān)th曠g景re第at刪er目c膛os碧t胞ef似fi乖ci桿en羽cy腎a洗lt荷ho蜂ug怎h遙th掏e線ef駱fe耳ct擊s枯we派re萬l鬧es微s片ev套id述en含t箏wh鞭en查f喊in巖an邊ci竿al希r跡at夠io訪s赤we緞re純u爐se吉d芳as賤t寸he區(qū)o櫻ut蓄co翅me堂m較ea旦su股re撿.村Pr姐as辰ad忠a格nd目H獨ar隨ke浴re智e凱xa重mi遭ne準d完th肉e善re家la染ti蜻on晉sh餓ip抹b床et橡we賺en輪t幸ec慚hn奶ol賄og腳y缺in怕ve慎st智me錯nt無a路nd念p桿er霜fo謀rm襖an琴ce瞎f醋or尿4池7枕re筋ta戲il銷b躁an糟ks灰a醫(yī)nd幼f斧ou磚nd砍p剪os淋it泡iv果e樸be薦ne齡fi熄ts界o怎f畜in禁ve刺st酸me起nt揭s姨in府I莖T脹st等af去f.

秒親W稿hi攜le孔t灶he寒se沃s供tu被di著es彩s回ho盈w養(yǎng)a務st件ro沉ng喪p加os亂it焰iv析e仰co宵nt戶ri島bu鏡ti養(yǎng)on拜o氏f戲IT安i腿nv尿es遮tm南en沈t著on扶a欠ve麻ra抵ge辯,藝th工ey贈d屢o液no歸t毀co采ns逗id寸er黎h敞ow故t定hi當s踐co復nt遙ri豐bu梢ti兼on琴(脫or場l卷ev肚el劉o分f劇in魄ve臺st攻me艙nt暫)耽va鑄ri鼓es博a礙cr語os燈s梯fi熟rm酷s.飛B雀ry紅nj皺ol柜fs波so夾n后an鞋d撤Hi揪tt啊f鍋ou普nd秘t租ha躬t丙"f旦ir姻m顆ef磨fe竿ct涌s"勿c螞an違a皺cc戶ou額nt旁f衫or牧a染s影mu衛(wèi)ch鋸a喘s絲ha慘lf哈t手he盡c寺on挪tr翁ib灶ut示io炒n蜻of餃I童T孕fo茅un榮d嘴in償t嘆he貫se胡e仰ar倍li纖er早s甘tu均di村es白.廉Re女ce洽nt成r濫es那ul躲ts僑s醉ug址ge漸st炭t辣ha使t告at作l邊ea圈st癥p鴨ar拿t編of日t襯he漏se習d趣if侍fe性re因nc枝es諷c加an孝b幫e涌ex問pl脂ai覺ne寄d竭by葉d腔if窯fe律re憑nc哄es議i鐮n值or窮ga炕ni潑za態(tài)ti駁on苦al娘a做nd越s城tr播at臘eg藥ic惹f遵ac串to幕rs淚.咽Br郵yn遣jo淚lf對ss勿on趣a熔nd衣H唱it奪t晚fo柿un醬d迅th零at舞f陸ir赴ms顛t途ha榨t謹us允e水gr浸ea喉te錯r也ov大er厘al離l叮IT膏b工en湯ef骨it扮s.晚B搏re揀sn熄eh男an莫,董Br賭yn匆jo慚lf晨ss眨on棒a組nd微H察it或t咬fo貫un凱d般a釋si灘mi皂la暖r姑re懷su聚lt狹f賀or部f漠ir棗ms晚t廊ha茂t聚ha摩ve閘g或re孟at窄er思l懸ev吐el腹s駛of寄s姨ki破ll卸s羞an叨d童th鍛os政e米th材at灘m烘ak磁e買gr耍ea換te南r緞in凱ve批st關me拴nt洞s善in樸t音ra農in晝in身g失an便d散pr東e-違em母pl默oy在me外nt梁s野cr需ee酷ni斯ng拼f量or煉h蒸um值an擾c添ap賞it涼al誰.勝I箭n屑ad盞di倦ti添on溫,活st強ra咐te靠gi球c戚fa冰ct葵or循s繡al坦so吵a看pp互ea假r團to臺a饑ff帶ec鑒t昆th腳e府va徒lu公e爭of星I帳T.臂F滴ir海ms茶t壟ha勵t望in剖ve擊st腹i誘n噸IT閘t赤o奔cr鉛ea知te竄c傅us光to緒me醉r乘va坑lu石e仗(e左.g猶.鑰im孟pr瑞ov芽e慧se壺rv籍ic似e,汽t接im付el忍in色es響s,務c臉on頂ve養(yǎng)ni稿en姑ce霸,稠va氏ri漠et卷y)該h尊av辮e漢gr寨ea夾te春r語pe蔽rf地or向ma倘nc雹e插th岸an閱f碼ir種ms定t蓋ha孝t耕in破ve譜st籠i價n壞IT姜t借o池re秧du辱ce覽c句os笨ts演.

競胖W落hi蘭le善t叛he蓋se平s仇tu挺di咳es盜a骨re介b污eg順in疲in乖g保to夸e喜xp腦lo秧re售h旱ow近t本he涂p奸er弄fo雙rm賢an食ce嫂o著f炮IT臂i計nv議es第tm聾en岸t捐va其ri錦es蕩a雄cr探os竄s等fi喊rm澆,革pa護rt廈ic英ul氣ar揪ly伏d漂ue吳t持o幫or筆ga裁ni圖za守ti凍on減al不a駁nd玻s溪tr撤at昂eg怠ic躺f涉ac選to刻rs歡,紐li瓜tt高le釋a滑tt償en梳t(yī)i地on肅h盈as局b秒ee勺n扇pa尺id磁t渾o澡th偵e摸te翻ch于no瘦lo挪gy倡d制ec兆is酷io誓n窯ma斃ki厲ng浙p撞ro疊ce倆ss昆.HowFinancialFirmsDecideonTechnology(PartFour)2.3ITInvestmentDecisions

Whilethereisnoconcisedefinitionof"bestpractice"inITinvestmentdecisions,thereareanumberofconsistentargumentsadvancedintheITmanagementliteraturethatcanbesynthesizedintoanunderstandingoftheconventionalwisdom.

ForthepruposesofdiscussionitisusefultosubdividetheprocessofITmanagementintosevendiscrete,butinterrelatedprocesses.Thefirstsixprocessesareorientedaroundtheproposal,developmentandmanagementofITprojects,whilethelastprocessisaboutmaintainingthecapabilitiesoftheITfunctionanditsinterrelationshipswiththerestofthebusiness:

1.IdentificationofITopportunities

2.Evaluatingopportunities

3.ApprovingITprojects

4.Themake-buydecision

5.ManagingITprojects

6.EvaluatingITprojects

7.ManageandDeveloptheITFunction

ThissubdivisionlooselycorrespondstomanyofthemajorissuesinITmanagementsuchasoutsourcing,linemanagement-ITalignment,softwareprojectmanagement,andevaluatingITinvestments.

Inaddition,thislistlooselycorrespondstoframeworksforthemanagementofIT.TheprimarydifferenceisthatthislistviewstheITmanagementprocessasmanagingastreamofprojectsratherthanfocusingonthefunctionoftheITdepartmentoverallortheroleoftheCIO,thetypicalperspectiveinthepreviousliterature.Forexample,acommonframeworkusedtoalignITtobusinessstarategy,thecriticalsuccessfactors(CSF)method,includethreeworkshops:thefirsttoidentifyandfocusobjectives,thesecondtodecideandprioritizeonsystemsinvestment,andthethirdtodevelop,deployandreevaluateprototypesystems.Boynton,JacobsandZmud(1992)identifyfivecriticalITmanagementprocesses:settingstrategicdirection,establishinginfrastructuresystems,scanningtechnology,transferringtechnologyanddevelopingsystems.Rockart,EarlandRoss(1996)proposeeightimperativesfortheITorganizationwhichcanbegroupedintomanagingtheIT-businessrelationship,buildingandmanagingsystemsandinfrastructure,managingvendors,andcreatingahighperformanceITorganization.Thus,whilepreviousworkhassubdividedtheprocessindifferentways,collectivelythestudiescoverallthesevenprocessesweexamine.

Wewilldiscusseachoftheindividualpointsindetailbelow.2.3.1IdentificantofOpportunities

Historically,theITfunctionwasprimarilyreactive,respondingtorequestsbybusinessunits.Abusinessunit.Abusinessunitmanagerwouldidentifyaneedforanewsystemorarepair/enhancementtoanexistingsystemandcommunicatethisneedtotheITfunction.TheITpersonnelwouldthenevaluatetheideafortechnicalfeasibilityanddevelopaprojectproposalincludeaninitialdeterminationofresourceneeds,cost,anddeliverytime.WhilethismakeseffectiveuseofITpersonnelinevaluatingparticularideas,itprovidesonlyalimitedroleforITpersonneltoaidintheidentificationoftechnology-basedbusinessopportunities.

Forthatreason,someauthorshavesuggestedthattheITfunctionshouldplayalargerroleintheidentificationoftechnologicalopportunities.Forexample,DavenportandShort(1990)emphasizethatITcapabilitiesshouldinformbusinessneedsaswellasthebusinessunitsplacingdemandsontheITfunction.Fockart,EarlandRossandBoynton,JacobsandZmudidentifytheroleof"technologyscanning"and"technologyeducation"asanimportantcomponentofacentralizedITdepartment;theyarguethatinformationsystemsspecialistsshouldbereponsibleforevalustingnewtechnologiesforbusinessapplicabilitysincebusinessunitswillgenerallylacktheresourcesorthetechnologicalcapabilitytoperformtheseevaluationsthemselves.Moreover,centralITisbestpositionedtoeducatetheendusestomakethemgood"custmers"ofthecentralITgroup.

Inthebankingindustry,ITmaybeabletoplayanadditionalroleincoordinatingtechnology.Becausebanksandotherfinancialfirmsareoftenmanagedwithlargelyautonomousbusinessunits(forexample,banksareoftendividedintoproductlines---cashmanagement,investment----oralongcustomersegments---wholesale,commercial,retail)onlythecentralITfunctionwillhaveaperspectiveovertheporfolioofsystemsprojectsandcapabilities.OnecriticalroleinthisrespectistheprovisionanddevelopmentofthesharedITinfrastructure(e.g.centralprocessors,networks,softwarestandards,etc.).OftentheseprojectsnaturallyspanbusinessunitssuchthattheonlyralowneristheITfunction;alsotheygenerallytendtobehighlytechnicalandthusthenaturalresponsibilitywouldalsofallontheITdepartment.HowFinancialFirmsDecideonTechnology(PartFive)2.3.2EvaluatingOppoutunities

Onceaprojectisatleastinitiallydefined,thereisaprocessbywhichtheinitialideaisconvertedintoaproposalthatcanbeevaluatedbymanagementforapprovalorrejectionoffunding.

Inthelasttenyears,ithasbecomemoreorlessstandardpracticestodevelopabusinesscaseorbusinessplanforanysubstantialITinvestment(somesmallmaintenanceprojectsaresimplydoneonrequest),althoughthecontent,sophisticationandformalityofthisprocessvariedsubstantially.Themosttypicaloftheseprojectproposals(assumingamid-sizetolargeproject)taketheformofabusinessplanwhichincludesaqualitiativedescriptionoftheobjectives,competitiveenvironment,adescriptionoftheopportunityand,insomecases,animplementationplan.Whiletheformoftheseplansvarieswidely,therearesomegeneralpointsofcomparison.

Forthequalitativeportion,themajorissueiswhethertheplanexplicitlyaddresseschangesinthebusinessenvironment,orisprimarilyinwardfocused.Forminorsystemsenhancementprojectswithnostrategicobjective(orevenmajorinvestmentsthatarenotstrategicsuchasyear2000repairs),itmakessensefortheplantofocusentirelyoninternalissues.However,totheextentthattheinvestmentismadeforcompetitivereasonsorislikelytospurareactionfromcompetitors,itisimportanttoqualitativelyevaluatewhetherthebusinessenvironmentwillremainstaticand,ifnot,examinepossiblescenariosthatarelikelytooccur.TheassumptionofastaticbusinessenvironmentisacommondecisionbiasthatcanparticularlyplaguestrategicITinvestments;Clemens(1991)termsthisthe"trapofthevanishingstatusquo".

Forthequantativefinancialevaluation,mostITevaluationmethodshavetheirrootsintraditionalcapitalbudgetingproceduressuchasdiscountedcashflowanalysis(DCF).However,whilethesetechniquescanworkwellforprojectswherecostsandbenefitsarwelldefined(e.g.purchasingoff-the-shelfsoftwareinpursuitofoperationalcostsavings),itisincreasinglyrecognizedthatsimpleapplicationofDCFapproachesisnotsufficientforITinvestments.ThisisbecausemuchofthevalueofmodernITinvestmentsislikelytobedifficulttoquantify--suchasrevenueenhancementsorcostsavingsthroughimprovedcustomerservice,productvariety,ortimeliness.Onecommonlyusedstategyistovaluenon-quantifiablebenefitsatzero,althoughthisstrategywillsystematicallybiasprojectevaluationstounnecessarilyrejectprojects.

RecognizingthelimitationoftheDCFapproach,severalalternativeapproacheshavebeenproposed.Onemethodistobasetghecaseentirelyonqualitativeanalysis;unfortunately,thisapproachoftenleadstohightlysubjectivejudgementsandislikelytoerronthesideofacceptingbadprojects.Kaplan,recognizingthisprobleminthecontextofevaluatingcomputerintegratedmanufacturing(CIM),proposedusingavariantofDCF;afirmcalculatesthepresentvalueoftheinvestmentusingallthecomponentsthatcanbequantifiedandthencomparesthisprliminaryvaluetothequalitativelistofotherbenefitsandcosts.Inothercases,wheretheevaluationismakedifficultbecauseoffutureuncertainties(e.g.marketgrowthandacceptance;responseofcompetitors)decisiontreesorothertypesofprobabilitybasedassessmenttoolsmaysimplifyinvestmentdecisios.Finally,forsometypesofinvestmentsordecisions(forexample,thedecisionwhethertoinvestimmediatelyordefer),advancedtechniquessuchasrealoptionscanbeapplied.

Althoughthisdiscussionhasfocusedprimarilyonevaluatingthebenefitpartofthequantitativeevaluation,thereareotherdifficultiesinestimatingthecostofITprojects,particualarlythoseinvolvingsoftwaredevelopment.ExistingmodelssuchasCOCOMOoffunctionpointsestimationareknowntoimprovetheabilitytopredictprojectlength,staffingrequirementsandtotalcosts,althoughtheyarekowntobesystmeaticallyoffbyasmuchas400%.However,theaccuracyoftheseestimatescanalsobeimprovedlaterintheprojectwhenspecificationsarewelldefinedorbycustomizingthemodelstotheexperienceofaparticularorganization.However,despitethefactthatthesetoolsandapproachesarereadilyavailablemanyfirmsstillutilize"seatofthepants"estimantsorlockinschedulesandcostestimantsbeforetheprojectsarefullydefined.HowFinancialFirmsDecideonTechnology(PartSix)2.3.3ApprovingITProjects

Onceaprojecthasbeenevaluatedandaformalprojectproposalexists,thereareavarietyofmechanismsthatareusedtodeterminewhichprojectsshouldbefunded.Mostinstitutionshavesomeformofcommitteestructure,similartocapitalbudgetingcommittees,whichisresponsibleforevaluating,modifyingandapprovingprojects.

MostofthepreviousliteratureonthemanagementofIThasfocusedontheso-called"ITsteeringcommittee"whichisanexecutivelevelgroup,oftencomprisedoftheheadsofbusinessunitsortheirdirectreports.TheobjectiveofthiscommitteeistoensurethatITstrategyisalignedwithbusinessunitstrategy,projectsarecoordinatedacrossbusinessunitswheretherearepossiblesynergiesandtoeducatethebusinessunitmanagersontheboththeactualactivitiesoftheITgroupandthepotentialITopportunities.2.3.4Make-BuyDecisions(Outsourcing)

Attheinceptionofanyproject,afirmhasthechoiceofwhethertoutilizetheirinternalresourcesintheITdepartment(insource)orutilizeanoutsidevendorforanyorallofaproject(outsourcing).Whilethemarketforoutsourced