It's Time to Take Control
- 05 August, 2002 09:57
CONSULTANTS: - Can't Live With 'Em, Can't Live Without 'Em.
Jerry Hale is giving the cold shoulder to the big consultancies that want to help him run his most important software projects. " I refuse to turn over the leadership of my ERP project to consultants," says Hale, vice president and CIO of Tennessee-based Eastman Chemical, which has 6000 users of SAP's R/3 ERP software. He felt burned by a consultant-led ERP project in the mid-90s that went over budget, over scope and far beyond the original project schedule. So when the time came to replace that system with a new one, Hale decided to use his own people to do the project along with a few handpicked consultants brought in to do small, specific tasks.
David Johns, senior vice president and CIO at Ohio-based Owens Corning, doesn't rely on consultants either. He and his IT staff now run Owens Corning's big software projects themselves, using consultants only for certain tasks.
Hale and Johns are not the only CIOs breaking with the way enterprise software projects were handled in the past - with armies of consultants from the so-called Big Five accounting firms. Many CIOs are now dividing these massive projects into smaller chunks, spreading them out over longer periods and either demanding more from, or doing without, the Big Five altogether.
CIOs are taking back control, in large part because they have no choice. These consultant-intensive megaprojects have a painful history of failure. In a Conference Board survey of ERP project managers released last year, 40 per cent of respondents said they failed to achieve their original business case even after being live for a year or more. More than 20 per cent shut down their projects before completion. For all companies, even the ones claiming success, costs were on average 25 per cent over budget and annual support costs went up by an average of 20 per cent over the supposedly inefficient jumble of legacy systems they replaced. Another survey, by consultancy Robbins-Gioia earlier this year, found that 51 per cent of companies were unhappy with the results of their ERP projects.
And now companies are beginning to see a replay of the ERP debacle with CRM. A recent survey by Fujitsu Consulting found that two-thirds of the companies that had installed CRM software failed to become any more " customer-centric" .
Though there are many reasons why projects fail, CIOs are now concluding that much of the fault lies with the pricing and delivery models of what used to be the Big Five accounting firms: Accenture (formerly part of Arthur Andersen), Deloitte Consulting, Ernst and Young (now Cap Gemini Ernst & Young), KPMG International and PwC Consulting. In a recent survey of IT and business leaders conducted by Peerstone Research in association with CIO (US), none of the Big Five rated better than a C in terms of the respondents' willingness to recommend them.
That is not good news for the Big Five. Their revenue growth has slowed in recent years (to 11 per cent, 8 per cent and 3 per cent growth for PwC Consulting, Deloitte Consulting and KPMG respectively in 2000), while IBM's IT consulting division, which trailed them all in 1999, now has nearly 50 per cent more consulting revenue than its nearest competitor, Cap Gemini Ernst & Young, according to Global IT Consulting Report. Cap Gemini Ernst & Young's growth came mostly from combining operations in 2000, according to the newsletter. Of the Big Five, only Accenture maintained the growth rate that the Big Five had come to expect in the 90s. Outsourcer EDS's consulting division, though much smaller than IBM's, is also gaining fast on the Big Five, according to the newsletter, growing by 72 per cent to $US650 million in IT consulting revenue. Their customers are happier too. IBM received an A-plus rating in recommendations in CIO's survey; EDS got a B-minus.
The resentment against the Big Five runs deep. Respondents to the Peerstone survey and IT executives interviewed by CIO claim the Big Five send them undertrained, underqualified people, and bill them at high rates. There is high turnover among these junior consultants, causing major disruptions in projects. CIOs also accused consultants of constantly adding in new work, blowing budgets and schedules. IT executives, it appears, are fed up with the way consultants get paid for every hour they put in on the project, regardless of whether the project is on time, on budget, or yields positive results.
With the Enron scandal forcing the Big Five accountancies to spin off their consulting divisions, the very survival of those divisions may be at stake. IBM is broadly diversified, EDS and Accenture are firmly rooted in outsourcing, but the rest of the former Big Five consultancies still rely on big engagements for a living. " There is a wholesale examination going on at these firms of the underlying principles that they operate under," says David Caruso, vice president and research fellow of Boston-based AMR Research. " That's the result of incredible push-back from users."
One thing is clear: the way these projects have gone in the past cannot continue if companies have any hope of getting a return on their vast investments - $US40 million to $US240 million for a typical ERP project in a Fortune 500 company, according to AMR Research. Even the consultancies agree.
" Western companies spent north of $US300 billion on ERP since the mid-90s, and if you put your hand on your heart, you can't say they got that back," says John Parkinson, chief technologist for the Americas for Paris-based Cap Gemini Ernst & Young, the second biggest IT consultancy in the US. " You look at it and say, I'm not sure that was the best thing to do - at least not the way it was done."
A Good Deal, but Not for You
In the early days of consulting, a few experienced consultants would sit down and noodle business strategies for their clients. They would charge high hourly rates and expenses - known as time and materials in the business - but only for a limited period of time. Then big IT came along, bringing enterprise software projects that were huge in scale, fuzzy in scope and seemingly endless. Armies of consultants were unleashed on Fortune 1000 companies to install the systems and redesign work methods.
Such contracts have become the golden apple for consultancies because they mean big deals, and big deals mean many consultants at the customer's site, all usually working on a time-and-materials basis. The archaic time-and-materials billing method does nothing to help control the scope or length of these projects. The consultants get paid for every hour they put in on the project, regardless of whether the project is successful or on time.
The Big Five depend on these enormous open-ended projects to preserve their profit margin, which typically runs anywhere from 20 per cent to 40 per cent or more. " Consulting firms make more money and get better margins by selling big deals," says Christine Ferrussi Ross, an analyst for Forrester Research (US). " The more people you have on one deal the more your cost of sale goes down. You're not out there having to look for another deal."
Consultants are also rewarded for selling follow-on work to customers so that once the company gets its foot in the door, it can stay there as long as possible. KPMG calls this its client-for-life strategy, according to Mark Lee, senior vice president of product solutions for KPMG (US). " Our intention is not to get into the client and stay forever on the project," he says. " It's building a strong relationship where the client looks to us to be a trusted adviser."
But customers have lost patience with the foot-in-the-door approach, according to Denny Wayson, vice president and chief analyst of enterprise solutions at Gartner Dataquest IT Services (US). " We don't think that vendors that use that approach are going to survive," he says flatly.
Worse, many of the consultants that arrive at a customer's door don't have the experience one would expect. " You get a mixed bag," says Craig Coggins, financial systems supervisor for Wacker Siltronic, a German semiconductor manufacturer, who worked with Deloitte Consulting on a recent SAP installation at Wacker Siltronic's US factory in Oregon. " You get some people who are very experienced and talented, and others who are intelligent, articulate people but don't really know the software and are learning as they go along."
Insiders call this the school bus effect. Born out of a drastic shortage of qualified people, the big consultancies began hiring young people fresh out of school. Though all big integrators hire youthful and less experienced consultants, Andersen Consulting (now Accenture) was best known for the practice. Eager to work the often egregiously long hours that these projects demand, the young consultants are trained quickly and armed with an explicit methodology - a sort of consultant's instruction manual - to help them get up to speed on enterprise software efforts.
Eastman Chemical's Hale says that inexperienced consultants were part of the reason his first ERP project was late and over budget. He felt the consultancy he hired was training its people on Eastman's time and dime.
Every big consultancy we spoke to for this story says it has scaled down the percentage of fresh-faced consultants assigned to enterprise implementations. According to James Hall, managing partner of technology and research for Accenture, which is headquartered in Bermuda, 70 per cent of its current consultants have prior project experience, and 30 per cent are junior-level people who may not have project experience.
The Yes Men
Even the most experienced and highly rated consultants have trouble delivering the changes in work methods that they promise at the beginning of these big enterprise software efforts. And when this transformation - promised first by the software vendors and seconded by the consultants - does not happen, customers are left embittered.
" We've invested $US200 million in SAP so far, and we don't have much to show for it," says Derek Dyer, manager of global e-business for Deere and Company, the Illinois-based farm equipment manufacturer based. Dyer doesn't blame the SAP software for Deere's lack of return on ERP. He singles out one of the consulting firms on the project, IBM Global Services, for criticism.
IBM Global Services was brought into Deere in 1999 to help make SAP the ERP standard across the company (some prior installations already existed). As part of the standardisation process, IBM was asked to help transform some of Deere's more antiquated work methods. But when Deere managers resisted the changes, IBM Global Services backed off, according to Dyer. " In some cases, we wound up mapping SAP to our current lousy manual processes, which made things even worse," he says.
In areas where the broken work methods were left in place, the new system performed worse than the rats' nest of old spreadsheets it replaced, says Dyer. " The managers ran the shop floor on an Excel spreadsheet before SAP, and if something didn't look right in the data, they'd just go in and change it," says Dyer. " Well, SAP doesn't let you do that. It's tightly integrated, and if you try to doctor data, it won't flow correctly through the system."
In November 2001, Deere halted its plans to roll out R/3 across the company and fired IBM. Deere's new plan is to install ERP only " where it is absolutely needed" , says Dyer. (IBM Global Services did not respond to repeated requests for comment.) All of the consultancies we spoke to for this story acknowledged that business process improvements are the weakest link in big software projects. But they say that's often because their clients don't believe they need it or don't want to pay for it.
" A lot of times the client just doesn't want to do it," says Jeff Balentine, global leader of Deloitte & Touche's technology, media and telecommunications group in Dallas. " They see it as touchy-feely stuff, and they don't want to pay for it." But these companies are often the ones that complain about their projects when they are done, he says. " They get it done fast and cheap, and then they say: Â'Where's the wow?' We as consultants need to stand our ground and say we need to do the process improvements."
But to improve processes, you have to reorganise the company, and that creates tremendous internal resistance. " These projects create winners and losers inside the company," says Dave Duray, global SAP leader at PwC Consulting, which is headquartered in New York City. " Everyone signs on and then when you start to deploy it, the individual managers see that their administrative staffs are going to be taken away and centralised and they say: Â'I don't want that.' That's when management has to step in and be tough, and I don't think they have been. At that point [consultants] should stop and say: Â'We're not going to do this. You're wasting your money.'" But consultants won't walk out. " If the client insists on something that's a bad idea, then we'll do it," acknowledges Cap Gemini's Parkinson. As a result, he admits, " I don't know of any engagement where we were allowed to do everything necessary to guarantee the client's success."
Time to Push Back
But Owens Corning's Johns does not believe consultants can ever successfully drive change in a client's company even if they're given the mandate to do so. Johns learned this lesson the hard way. In 1997, less than a year into its SAP project, Owens Corning renegotiated the contract with Deloitte Consulting to take back control of the project. Owens Corning employees didn't want to listen to Deloitte. They wanted their bosses to tell them what to do, Johns says. So the new contract stated that project responsibility remained with Owens Corning and specified the roles and responsibilities of all consultants on the job and what they would be paid. Instead of paying them on a pure time and materials basis, the new contract stipulated that the consultants would get their hourly fees only after they delivered specific pieces of the project by certain dates. " The time payments got everyone's attention," says Johns.
The issues of cost and pricing create the most distrust between consultants and their clients. Wayne Hall, CIO of The Nautilus Group, the manufacturer of Bowflex, Nautilus, Schwinn and Stairmaster health and fitness products, sent out a detailed RFP for his ERP project and hired all five consultancies that responded, asking them each to do small projects that would feed the eventual ERP project. Despite spending months at The Nautilus Group getting to know the company and its processes, all five came in with bids of $US3 million to $5 million, less than half of what Hall expected.
So Hall sent the RFP back to one company, Emerald Solutions, along with more information and detailed specifications. Emerald revised its bid to about $US5 million. Then, after he hired Emerald, the estimate ballooned to $US15 million. Hall believes the consultants purposely lowballed the initial estimates to get the job.
His CEO took one look at Emerald's bid, Hall says, cancelled the contract and turned to Hall and said: " Wayne, you do it." Hall had to rebuild his IT department to staff the ERP project. He hired Deloitte (which also submitted a $US3 million initial bid) to work on specific slices of the project.
Breaking the Consultant Habit
How can CIOs survive the enterprise software party without bad hangovers? Some major trends are emerging. It's clear that IT departments must build up their own project management capabilities. A strong internal project management capability acts as a check against the natural tendencies of the consultancies to grow projects and add new ones. In the Robbins-Gioia survey of companies that had installed ERP systems, those who had created in-house project management offices were successful 15 per cent more of the time than those that did not.
Good internal project management is even more important in the current economy, when companies are scaling back their software projects and splitting them into smaller pieces. If the consultants work on small pieces of the project, the CIO must make sure the pieces all fit into an overall business strategy.
For CIOs who cannot afford to build an IT department to run an enterprise software project, pricing schemes provide the leverage that can keep their consultants in line, on time and under budget. Devices such as fixed-fee, revenue sharing and not-to-exceed limits act like invisible project managers.
In fixed-fee engagements, consultants are paid a specific fee negotiated before the project begins. A recent Gartner survey reports that 62 per cent of consulting customers called this their preferred strategy. But consultants don't like fixed-fee projects and will want to place rigid controls on project scope to make sure they can still make a profit.
Such fixed-price contracts can work if CIOs clearly define the project up front, with a tightly fixed scope and set of deliverables. On big, unpredictable enterprise software projects, that is next to impossible to do. Big fixed-fee projects can easily dissolve into endless bickering over " change orders" , time delays and scope changes. Consultants will also be tempted to cut corners if they're falling behind. Fixed price works best on small projects or on big projects that are broken up into small, clearly defined chunks.
Consultants, of course, much prefer time and materials, saying it lets customers change scope or goals when necessary without rewriting the contract. However, customer backlash has driven consultancies to come up with new pricing models to share risks and rewards with clients. One is agreeing to a lower hourly fee in return for sharing a piece of the cost savings it delivers to customers. Twenty-five per cent of Gartner's consulting customer respondents said they would consider some form of shared risk and reward model, but just 4 per cent of those surveyed said it was their preferred strategy.
To keep project costs from spinning out of control and yet maintain enough flexibility to achieve business process changes, a mix of time and materials and fixed cost makes the most sense. Time and materials with a not-to-exceed cap or time and materials with a contractual schedule for delivery of benefits are two popular options. For example, when Joseph Wolfgram, former CIO for the Bill and Melinda Gates Foundation in Seattle, hired a consultancy to help upgrade the Foundation to Windows 2000, he agreed to time and materials, but inserted performance milestones in the contract that specified the number of users that would be up and running by a particular date. If the consultants missed that date, they agreed to a reduced hourly rate until they got it done.
With the era of huge, open-ended enterprise software projects at an end, the Big Five will need to develop ways to make money on smaller projects while delivering the ROI their customers expect from enterprise software. It is a tall order. They have to deal with a legacy of mistrust from the first wave of big enterprise software projects, and a flat economy is making their customers more cautious and demanding than ever. If the Big Five can't develop new pricing and delivery models to drive the next phase of the enterprise software revolution - supply chain collaboration with customers and suppliers - they may not survive.