Questions to: Thomas Schmidt
about cost planning for SAP application development
Thomas Schmidt is Manager of Performance Measurement at HVB Systems GmbH
“The whole thing can only be handled with the help of benchmarking.”
Mr. Schmidt, HVB Systems is generally regarded as a very good company. Nevertheless you have identified areas that require improvement and you have evaluated your application development procedures in the SAP sector. Why?
Schmidt: The key word is transparency. We wanted to know more precisely how productive our application development projects actually are. On the one hand, our management also uses productivity figures to control target agreements using the balance scorecard; on the other hand, we at the HVB Group tell our customers how competitive we are and that no external supplier will be able to offer better value for money.
In other words, you wanted to show how good HVB Systems actually is?
Schmidt: That was one point. We wanted to know how the productivity of our application development had developed over time and how we compare with other suppliers. The whole thing can only be handled with the help of benchmarking; our peer groups were selected very carefully and included not only IT suppliers — of course — but also IT units run by other financial institutions. But we were not just interested in establishing a location or (another important point) permanently acquiring actionable ideas to improve our processes. Above all, we wanted to use the parameters gleaned from the process to enable us to estimate the costs of future application development projects better — before we actually start the projects.
But surely we are always being told that this is not possible — that SAP application development cannot be planned?
Schmidt: That is certainly the claim. But with the help of Maturity and its benchmark data, we managed to develop a measurement method that allows us to model productivity and evaluate it. And thanks to the parameters we generated from this process, we can also cost the expenditure levels of future application development projects much more accurately.
How does that work?
Schmidt: For projects of a certain size, the project leaders have to plan the expected expenditure in so-called initial studies. At the same time, we produce a second, independent expenditure estimate. The basis of this is a questionnaire that the project leaders have to complete containing questions about all the risk and context factors that may influence performance that can be identified at this early stage. The specialist requirements are analyzed in detail and their complexity is assessed. Using the complexity and expenditure parameters and the benchmarks from the measurement method, we calculate the probable expenditure. The advisory meeting that is then held with the project leader forms the important core of the process, helps avoid errors, and optimizes expenditure planning. And naturally this also creates a certain amount of commitment to compare oneself with the normal market standards.
Does this not remove some of the power from the project leader?
Schmidt: No, not at all. He is responsible for the project and therefore he must make the decisions about, for example, the actual level of the quote to our customers. On the contrary, our project leaders are very grateful for the support, because we can give them additional security in their efforts to calculate a price that is attractive to the customer, meets the requirements, and is commensurate with the service being provided. All of our customers now naturally want to know in great detail what they are going to get and how much they are going to have to spend on it.
And what has been the overall effect of the process?
Schmidt: We compared SAP projects that were completed in 2003 and 2004 with each other and were able to identify a very good increase in productivity within our own company and compared with the benchmarks. In fact, this increase was well above the targets that we had set ourselves on the basis of the balanced scorecard. We are now able to accurately analyze the efficiency of our projects — and thus also the success of the many process enhancements that we are permanently initiating and implementing. There was another important effect — we can use the expenditure estimate to prevent expensive failures. But, on the other hand, we now also have the transparency to identify positive project examples and to learn from them.
Download the interview IT benchmark at HVB Systems as a PDF file