When did you last see your outsourced service partner?

It’s not what you do, it’s the way that you do it. And so it is with outsourcing information technology services. Research from both the US and the UK suggests that even though suppliers may meet agreed service levels, customer dissatisfaction remains. It’s not the outsourcing itself that determines business performance, but the way that outsourcing deal is negotiated, defined, incentivised and managed. A major survey published in April by US-based strategic IT consultancy Technology & Business Indicators found that every single one of its respondents would continue outsourcing IT functions. “That surprised me,” says Stan Goldman, TBI’s president and chief executive officer, “because a lot of these deals had dissatisfaction associated with them.” The respondents included big name organisations such as Kodak and together accounted for around 6% of the IT outsourcing spend in the US. The survey found that 59% of respondents perceived outsourcing positively overall, leaving a sizable 41% that did not. There were several causes of dissatisfaction, but contractual breaches were not among them. “Outsourcers were seen as delivering on their contractual guarantees and on contractually agreed service levels,” says Goldman. However, clients were less impressed with outsourcers’ performance in the areas of product development and the management of distributed environments. The TBI survey also highlighted what appears to be a recurring theme in the world of IT outsourcing – the problem of tying the service into business strategy. “In almost all cases outsourcers were seen as not delivering on providing business strategy,” says Goldman. In the UK, a survey by KPMG last year found similar determination to stick with outsourcing, despite some grievances over the service. The survey, The Maturing of Outsourcing, was based on 123 UK organisations with experience of outsourcing IS or IT. Although 90% of respondents were generally satisfied with the level of the outsourcing service they received, three-quarters expressed significant dissatisfaction with at least one important aspect of the deal. Where buyers used the supplier’s contract, they were less likely to feel they received a good service and were less likely to feel they had the flexibility they wanted. The feeling of dissatisfaction was three times greater if adequate service level agreements (SLAs) were not specified in the contract. Clients who don’t look after their own interests by negotiating at least basic SLAs have only themselves to blame, since there is now much more guidance available and both clients and suppliers are more experienced. KPMG’s guide to SLAs advises that the SLA should identify what the service deliverables are, when they are to be delivered and where. For each service deliverable the SLA should define: the agreed service measures and levels, such as availability and response time; the boundary within which the service level is agreed (for example, the workload); and the price charged for the service. The SLA should not describe how the service is to be delivered. Leaving out the ‘how’ is something IT people often find hard to do. “If IT people are involved they want to know the underlying bits and bolts and boxes,” says Martin Burvill, business manager at BT Syncordia Solutions. “But that is substantially irrelevant. What you are buying is a service.” What matters are items such as the time taken to get a service back on line, or the level of access to a facility. Richard Jones of solicitors Clifford Chance also backs up the wisdom of not drafting contracts in ways that are technologically dependent, that is, specifying the type of computer that will be placed on staff desks. If you name the model of PC you intend to use, then every time you upgrade, the contract has to be redrafted. “IT specialists tend to find it difficult to make that jump,” says Jones. IT outsourcing contracts must be kept flexible. “The industry is moving so quickly that you don’t know what your requirements will be down the line. You need to draft the contract as simply and generically as possible so that you can make changes.” That said, Jones adds that contracts should be specific about the customer’s requirements, with the consequences for failing to meet them spelt out clearly. “The typical remedies would be service credits – fixed reductions in the fees if service levels are missed by specific amounts.” Contracts could also allow for the ultimate sanction – the right to terminate the contract if the service is extremely poor. In general, most contracts do now include an SLA. The question is whether these SLAs are good enough. “SLAs are not in themselves sufficient for the types of contract people want today,” says Bob Aylott, principal consultant in charge of outsourcing at KPMG Management Consulting. “They are necessary, but not sufficient. If you are looking beyond process control, and asking what does this contribute to my business, SLAs do not do that.” KPMG is looking at key performance indicators (KPIs) as a means of getting round this problem. “Even in the IT area, people are saying they want to incentivise the supplier on the basis of end user satisfaction as a KPI,” says Aylott. “We are incentivising suppliers on the basis of KPIs, and penalising them on the basis of SLAs.” Dan Cohen, managing principal for outsourcing at IBM, believes that the key to getting the service delivery right is to find a way to share not only the risks associated with the deal, but also the potential for benefits. He says contracts in the past have been too focused on shifting risk. “We try to share the risk and the reward,” asserts Cohen. IBM’s approach also tries to build in ‘transformation’, essentially the introduction of greater automation and the reduction of labour. “Labour rates are going to go through the roof,” Cohen says, referring to the Year 2000 issue. “The way you get to better service levels at a better price is by investing in transformation.” The appropriateness of applying business-performance measures to outsourcing agreements clearly depends on the potential for the service provider to make an impact on that particular business function. The more involved the service provider gets in developing strategic processes and services, then the more relevant it is to measure primary business indicators, says Duncan Aitchison, executive director of Cap Gemini. “We will always look to see if we can track ourselves in ways as closely linked to the customer’s business as possible.” For example, if devising help desk systems, cost per enquiry could be a useful measure. “But if you have no leverage over the outcome, then it’s probably unreasonable to use such a measure,” Aitchison warns. He stresses that it is not enough simply to draw up an SLA at the start of a contract and then leave it. “Without continually thinking about whether the levels measured are correct, those levels could be being met but the customers could still not be getting what they want,” he says. “Cap Gemini tries to keep tracks on what its customers think through satisfaction surveys, the aim being to highlight if there is a problem emerging and then go and have a dialogue,” Aitchison says. “It’s an early warning device.” BT’s Burvill believes that absence of input from top-level management is often the reason why outsourcing has failed to deliver results in line with expectations. Top managers may take the initial decision to outsource, then leave the implementation to lower-level, less empowered managers. “It works well where senior management say they will do the negotiating, rather than push it onto someone else,” Burvill says. End-users also need to be involved in negotiating the service to be delivered. “Often the trade-offs have to be made between the network and applications people and end-users,” says Burvill. “You need someone to bring the interest groups together.” To do that effectively may require input from someone with significant stature, perhaps the chief operating officer or chief executive. What everyone has to keep in mind is the end goal – the final, total service aim. “The industry as a whole has gone a little astray with SLAs, in that the solution for the end user, the client, ends up getting chopped into bits,” says Gareth Cadwallader, director of the UK’s multivendor customer services arm of Digital Equipment Corporation. “Each bit gets measured but what gets lost is the overall solution that the client receives.” Cadwallader believes clients care about two key outcomes – service availability, and the speed with which problems are solved. SLAs which are being met can still mask causes of significant annoyance to clients, he believes. Say, for example, that the SLA states that helpdesk queries will be handled within 30 minutes for 90% of the time. The outsourcer could meet that target, but one or two “horror stories”, with calls left unanswered, could cause major client dissatisfaction. Dealing with these would make little difference to the percentage measure of performance, but could make a huge difference to the client. “The biggest challenge to the SLA industry today is doing the hard work of improving the way SLAs improve processes as seen by the end user,” says Cadwallader. One of the problems faced by those involved in finalising service levels is how to assess the impact on the business of different service elements failing to perform. One way round this is to model the IT system using a tool such as Observant, developed by Tertio to monitor SLAs. “We can model all the aspects of a local area network, for example, and show how something that goes wrong could impact on the end-user,” says Martin Hopkins, project director for Observant. As an example, Hopkins points to City trading systems. “We can show what it means to traders if a particular link goes down,” he says. Observant essentially translates the effect of an IT failure into a business repercussion. “If I am driving a car I don’t really care whether all the tappets are working or not,” says Hopkins. “I just want to know if the car needs a service.” If outsourcing contracts do not always achieve high customer satisfaction scores, few are in fact terminated. Nor is it common for clients to bring the IT service they have outsourced back in house. This appears surprising, given that many clients believe outsourcing fails to deliver on a business level. So why do so many companies stick with their outsourcers? One key reason is simply that, having removed the infrastructures needed to support an in-house service, they are unwilling to incur the cost of re-instating them. Referring to his survey group, Goldman says: “All these guys wanted to avoid capital expenditures. Cost savings declined in importance over time to CEOs, but cost predictability was extraordinarily high in importance.” Service providers back up that reading of clients’ concerns. “Our view is that, when we go to a client, most of them are interested in cost, even if it’s just delivering predictable cost, not reduced cost,” says IBM’s Cohen. This leads to the concern that clients could find themselves effectively locked into contracts that fail to deliver what they really want. “The danger of getting locked in is very real,” says Aylott. “Even if you write a contract to facilitate change, it is still difficult. If you look at renewal rates, there are a lot more renewals than people’s satisfaction statistics would lead you to expect.” Aylott says that transferring a contract to another provider, or bringing the service back in-house, is always difficult. “But it is near impossible if it is not thought through up front,” he warns. “You need leverage, and negotiation before the deal is signed is the only time you have that leverage.” Experience shows that to get the best from an outsourcing deal, both in terms of getting the service you want that delivers business benefits, and in terms of making an exit if need be, you need to put in the effort before the contract is signed. Only then can you expect your outsourcing deal to support you in the business of doing business.

 Causes of dissatisfaction with IT outsourcing deals >----------------------------------------------------------------R Over-dependence on supplier                                    42% Locked in to supplier                                          39% Lack of influence on the service levels of the supplier        37% Length of time in getting the service right                    37% Limited or no control over supplier                            33% More management time spent dealing with supplier issues        33% Loss of skills on IT applications                              32% Difficulty reverting to in-house service                       31% No longer in control of logistical resources                   29% Complacency of supplier service                                28% Increased operating costs                                      22% >-------!-------!-------!-------!-------!-------!-------!-------!--------R Source: The Maturing of Outsourcing, KPMG Management Consulting.