Organisations are under pressure to increase the cost efficiency of their IT environments and for some time forward-thinking CIOs have been ramping up benchmarking activity to measure the cost/performance of their IT infrastructures and support services.

In order to do this, they need to be able to measure where they stand against enterprises of a similar size and complexity, and against best-price industry norms. One of the questions implicit in benchmarking is whether it is more efficient to support IT in-house or to outsource some or all services. And if IT is already being delivered by an external service provider, are you paying competitive rates and getting a quality service? Benchmarking against up-to-date peer and market data can provide comparisons, indicate where money can be saved and can negotiate with suppliers on the client's behalf.

Benchmarking projects have tended to focus on various elements of the IT infrastructure such as back office, networks and desktop support, but a new area that is gaining traction is benchmarking applications such as ERP. Given the complexity of most ERP environments, some may say that trying to quantify it is like asking 'how long is a piece of string?'. However, identifying even small efficiencies can result in substantial savings. Licensing costs alone of a typical SAP implementation can account for up to 20 per cent of the overall ERP budget. Add related infrastructure elements such as database software, networks, storage, hardware and so on and the costs of running an ERP environment can vary widely depending on operational efficiencies, legacy components and interfacing complexities.

In a recent study undertaken in Germany, consultants from Metri Group compared the cost structures of several external SAP support providers. Before undertaking the study it was necessary to ensure that all of the client enterprise environments had similar levels of enterprise-critical relevance, usage intensity and volume, database size, technical complexity and service levels so that we could make a like-for-like comparison. Where there were differences, these had to be evened out to achieve a fair comparison.

Having set this standard benchmark, measurements were made to identify the range whereby an enterprise and its service provider deviated from the norm. Among other things, the study looked at service provider costs for incident and problem management, user administration and change management. The results revealed that outsourcer costs for the technical application operation ranged from €21,000 to €27,000 a year per SAP module. Given that this study involved only a few of the components that make up an SAP environment, it gives an indication of the magnitude of savings that can be achieved using application benchmarking.

While this case offers a glimpse into outsourcer pricing differentials, a comprehensive best-practice pricing survey must start by looking at a company's own cost drivers. The most important place to start is to determine how critical the SAP application is to the enterprise as a whole. This question can best be answered by quantifying the consequences of an SAP outage and then weighing up whether the level of mission-criticality justifies the costs of a higher level of service. Since most SAP applications are designated as highly mission-critical they typically involve premium SLAs including 365/24/7 support, which means there is usually very little in the way of cost reductions to be found.

The next cost driver is the range of services to be included in the service provider contract. These range from basic services such as process monitoring, incident resolution and ongoing systems maintenance to job scheduling, capacity management and technical innovation. There is more room to manoeuvre here as some service components may be retained in-house.

The third factor that determines cost is the complexity of the SAP installation. This includes the number of interfaces to other applications, the number of users and the number and distribution of enterprise locations both locally and globally.

Clearly, the more complex the environment and the greater the number of touch points, the higher the support costs. For some multinational firms, up to 70 per cent of their SAP service requirements may involve international locations. Rationalising this into regional service hubs may offer major cost and performance efficiencies but, given the investment in change management, this should be benchmarked first to identify exactly where, how and how much cost efficiency can be gained. Of course, Like any other aspect of IT, application benchmarking is an investment that must have a compelling business case. Our experience shows that an effective application benchmarking project can result in double-digit percentage savings for an enterprise that spends £500,000 or more per annum on ERP support.

About the author

Don Rekko is managing director at benchmarking, sourcing and governance consultants Metri Group