Quelle: Darwin, USA
In a teleconference last spring with Wall Street analysts, NikeChairman Phil Knight aired his frustrations with an underperformingsupply chain management system, which he blamed in part for Nike'slower-than-expected financial results. Asked Knight rhetorically,"This is what we get for $400 million?"
Like his shareholders, Knight wasn't happy. And he had an inkling why.Because the costly supply chain system wasn't tied in to other keyNike technologies, its anticipated business benefits fell criticallyshort of the mark. Without access to important information lockedwithin these isolated systems, the supply chain project's usefulnesswas limited.
Nike isn't alone in expressing disappointment with technology'spayoff. There is considerable unhappiness in many executive suites.Perhaps you, too, are wondering what happened after your company spentall that money to acquire new and strategic capabilities in the areasof e-business, supply chain management and customer relationshipmanagement (CRM), to name a few.
How did this epidemic of discontent come about? Some recent historymight offer a little context.
The year 2000, in particular, saw an unprecedented technology buyingspree, triggered in part by the mandatory diversion, during 1998 and'99, of IT investment priorities and dollars toward Y2K remediation.Once the millennium rolled over more or less without serious mishap,business emerged from its bunkers to deal with a lot of pent-updemand. It was time to get cracking on those long-delayedprojects - along with some fresh competitive realities. Juicing up thealready high sense of urgency was a brand-new business anxiety: "Howare we gonna keep from getting Amazoned?" A pervasive sense of needingto spend fast and furiously to catch up with the Internet phenomenondrove a period of frantic buying.
Naturally, those who signed the big, scary checks expected that bingebuying would soon be followed by impressive competitive gains. But intheir haste to get someplace fast, many companies neglected the messyreality that their back-end infrastructures consisted of a stew ofincompatible systems that could not usefully interact with oneanother.
Consequently, many newly purchased applications failed to deliver thepromised benefits, says Tim Talbot, senior vice president ofinformation technology services at PHH Arval, a vehicle fleet-leasingand -management company headquartered in Hunt Valley, Md. "Executivesgot sold a product that was supposed to do all these things butcouldn't do them all if systems integration wasn't [in place]. A CRMapplication is not going to stand alone, but businesspeople don'tnecessarily understand that," he says.
In fact, CRM and various e-business systems demand more than basicintegration, which means hooking up two systems so that they cancommunicate. To get real business payback from these applications,companies must make sure that the entire technical infrastructure isinterconnected so that business information can flow freely throughoutthe enterprise.
Wouldn't Drano Work Just as Well?
Attaining that free flow of business information is more than atechnical exercise in linkage; companies that get it right can usethis new ability to completely rethink the way business is done,maximizing competitive value by offering heretofore un-heard-ofcustomer benefits and services.
Look at PHH. The company has been integrating its systems since 1995,when it gambled on an application called PHH Interactive. The systempulls information from dozens of once-separate PHH applications andassembles a unified mother lode of data. As a result, thousands offleet managers at PHH's customer facilities can freely accessinformation as varied as car maintenance histories and trends, driveraccident rates, fuel transactions, repair-cost comparisons acrossvarious automobile makes and models, and billing information. Thisgives the fleet managers an opportunity to optimize their resourcesand control expenses - making them incalculably grateful to PHH andthus less likely to bolt to a competitor. Indeed, since implementingPHH Interactive, the company has been able to cut costs by 30 percentwhile signing on 20 percent more business; it is the number-two fleetmanagement company in the country, behind GE Capital's fleetmanagement division. The company considers PHH Interactive to be sucha competitive advantage that it often sends IT staffers out on salescalls to demonstrate the application.
Then there's Dell Computer. Considered one of the poster children forgood integration, the company has tied together not only its ownsystems but those of its suppliers. Called Valuechain.dell.com, theintegrated system has kept Dell's inventory down to an incredible fourdays' worth of supplies (compared with its competitors' 30 to 50days). Dell's ability to use customized websites to feed its biggestcustomers such vital data as order-tracking and billing informationwould be impossible without full integration of its disparate systems.In part because Dell has driven inefficiencies out of its supply chainand can interact directly with customers, it has become the number-onePC maker in the world.
OK, so Where's the Catch?
Such examples may be alluring, but they are not easy to achieve. Agreat many companies have balked at the idea of fully integratingcorporate systems for a couple of very basic reasons: Integration is atall - and expensive - order. Forexample, CTO Julie Hall recently dealt with a major systemsintegration project when her company, Advance Paradigm, in Irving,Texas, merged with PCS Health Systems to create AdvancePCS. Hallcounted several hundred disparate systems between the two companies.Her integration project timetable: one year. The cost: 50 percent ofher total IT budget.
Fifty percent of an IT budget is a lot of dollars but not an outsizeamount when it comes to integration. Beth Gold-Bernstein, vicepresident of strategic services at Ebizq, an integration informationportal based in White Plains, N.Y., estimates that a large integrationproject will cost between $500,000 and $1 million for software andlicensing, and implementation costs will run twice as high.
But in spite of the cost and difficulty of integration, the vision ofreaping Dell-like benefits ought to make such projects a top corporatepriority, says Bob Parker, vice president of e-commerce strategy atAMR Research in Boston. Yet despite their importance, he says, manyintegration projects fail or get abandoned. Why?
It's tempting (and all too easy) to line up the usual whipping boys oftechnology failure: overhyped software, unresponsive IS staffs,incompetent systems integration firms. But Frances Karamouzis, aresearch director at Gartner in Stamford, Conn., says that theproblems don't lie with technology alone. In fact, she says,"Technology is usually ahead of the curve from the perspective ofintegration." The real culprit, as she sees it, is a failure ofexecutive leadership. In other words, look in the mirror."One reasonmany companies are not well integrated is the amount of changerequired on the business process side, in addition to the technologycomponent," says Karamouzis.
Most business executives mistakenly think of integration as theexclusive province of gearheads: The techies will fix it down in thebowels of the engine room. But true integration also requiressignificant involvement from the business side of a company, andthat's not happening at many places.
"Businesspeople want the magic of what integration is going to do, butthey're almost always unwilling to pay the price in terms of what it'sgoing to cost them in extra work," says Steve Morelli, CFO at StarKistFoods in Pittsburgh.
In fact, Morelli says, "[That resistance] is a big part of the reasonthese projects fail. Businesspeople get into implementation, getoverwhelmed and either fight it or give up on it."
"I think, for the most part, businesspeople can appreciate the amountof effort that has to go into integration. But let's face it, theydon't really understand the guts of it," agrees Linda Reino, CIO atUniversal Health Services (UHS), a $2.6 billion hospital-managementcompany in King of Prussia, Pa.
Yet getting businesspeople on board integration projects is a criticalissue, says PHH's Talbot. "If you don't get the business aspect ofintegration nailed, you may as well just stop."
Nailing that business aspect, however, means managing the changeswrought by integration, and that's a huge up-front task lackinginstant gratification, says Gartner's Karamouzis. Just one example: Asintegration links processes across departments and divisions,executive sponsors will need to gird themselves for the turfskirmishes that inevitably flare when the walls between departmentscrumble.
Sounds hard? It is. But there are things business executives can do tomaximize the chances of success for integration projects.
So, We Can Just Integrate What Makes Sense?
Start by analyzing your company, advises Lisa Reisman, director of thedigital strategy and research group at Andersen Business Consulting inChicago. According to a recent study by Reisman and fellow analyst KimCollins, different types of companies have varying integration needs.For instance, if your company doesn't drool over leading-edgetechnology, it might make more sense to choose software from a bigenterprise software vendor, such as SAP or PeopleSoft. That's becausetechnology laggards can typically afford to wait for such vendors toroll out add-on software modules for different business functions. Theintegration benefit is that a single vendor's add-ons are easier toglue together. Conversely, companies whose competitive edge is honedon the latest and greatest technology are more likely to addcapability in such areas as supply chain management or CRM from theso-called best-of-breed vendors, who tend to get their products tomarket earlier than the big enterprise software companies.
David Root, the CFO at Eagle's Flight, a training company based inGuelph, Ontario, suggests following up with some cost-benefitanalyses. "Integration is an ROI decision," he says. "You need to askwhat's the dollar cost, what's the cost in terms of time and people,and what's the relative benefit?" The outcome will help executivespinpoint which business processes merit integration, he says. Noteverything does, as crunching the numbers can show. "Lots of times thecosts of integration are a lot higher than people would like, and thebenefits a lot less," says Morelli. "It doesn't mean it's not theright thing to do, but it does need to be thoroughly lookedat."
PHH's Talbot says that his IT group works closely with businessexecutives to identify areas that are ripe for integration. Theycompare the current costs of managing the process with the estimatedcost of integration.
"For example, let's say we use a certain up-fitter to customize pickuptrucks with bucket lifts," says Talbot. "Right now, the process isdone by phone or fax and takes about two hours of a PHH employee'stime. But we only need to have that kind of work done maybe once ayear. So even if it would only cost $5,000 to $10,000 to integrate oursystems with those of the up-fitter, it doesn't make sense. If we didtwo or three such jobs a month, it'd be a different story."
Talbot takes it one step further, suggesting that cost-benefitanalysis should drive the actual level of integration. There aredifferent ways of integrating systems, each with corresponding prosand cons. A batched integration process, for example, means that twodifferent systems send updated batches of information to each other atpreset intervals - every 24 hours, perhaps. Although this type ofintegration doesn't provide continuous updates 24 hours a day, it doeshave the benefit of being relatively inexpensive and easy toimplement. Real-time integration, on the other hand, links systems soclosely that information is continuously updated. While this hasobvious business benefits, it's also more complicated and expensive toachieve.
The trick is letting the needs of the business dictate the integrationstrategy. The radiology systems at UHS's hospitals are integrated inreal-time - though only to a point. "If you need an X ray, the orderhad better be there 10 seconds after you press the button," says CIOReino. But hospital charges are sent to the hospital billing systemonly at the end of the day. "Hospitals don't need to presentpoint-of-service bills like retailers do," she points out. "So whyshould I maintain that kind of overhead?"
But I Like to Badger the CIO!
Business executives should also respect the technical issues of anintegration project as well as the business drivers. Say, for example,that the vice president of sales is pushing the CIO to get a CRMproject done fast. The technology executive has to decide betweenmeeting the sales vice president's deadline or completing theintegration piece of the project in the most thorough way possible.It's a tough choice to have to make.
"The CIO will try to push back," theorizes StarKist's Morelli. "But atsome point he'll realize, 'Here's the momentum of the organization. Ican either get with it or get run over by it." A tremendous amount ofpressure is put on CIOs, Morelli says. "Most business users want thisstuff because they're at their wit's end and their bosses are pushingthem for more improvement." The best the CIO can hope for, he notes,is to manage the process so that people understand and buy in to theimportance of integration as the project unfolds.
Once integration is under way, executive sponsors should help businessusers understand the degree of change that the various projects willbring to their work processes. As different departmental systems aretied together via integration, business processes are too.Consequently, newly linked tasks can require a heretofore unknowndegree of cooperation between and among departments.
Morelli gives the hypothetical example of a purchasing department'sbusiness process. Pre-integration, it might take employees sixdifferent tasks to complete a purchase order. But once that isolatedpurchasing system becomes integrated with other corporate systems, theunified purchasing process might require 60 tasks. Those new tasks maywell mean that the company can reduce order errors that result fromrekeying data, or keep inventory levels low. But for users faced withnew tasks, sometimes it's hard to see the forest for the trees, saysMorelli. "Business users don't recognize that integration is going toimpose a new level of discipline and workload that they don't have todeal with today."
But he insists that doing so is vital. "No matter how integrated thesystems are, you won't get things to work unless the informationtherein is also integrated. The interrelated aspects of the businesshave to be kept up-to-date," he says.
Here's the hard truth: Integration has been an enormous challenge foryears. With the growing complexity of technology environments, coupledwith escalating business pressures, it won't soon get any simpler. Butdespite the difficulties, the potential payoffs are overwhelming. ForJim Trotman, director of strategic development at PHH, the resultshave been a revelation. Over the years, he says, "I've had times whenI thought of integration and architecture as a pain in the rear. Butnow that I've lived in a world where it gives payback month aftermonth after month, I've almost become an evangelist for it." (Notethat ambivalent "almost.")