Why government run shared services often fail, part 2

This blog is a follow up on this piece in which I wrote on some of my experiences with IT shared services in the government sector. Here I want to write about two solutions which would improve the success rate of a shared service center.

Practical example

The organization is a grouping of local government agencies which have a long history of semi independence. Policy is partially dictated from a ministry and national politics, but there is also room for local decision making. To align, the governors of the local agencies gather on a yearly basis to agree on long term policy and cooperation.

One of the decision was starting an IT-shared service center some ten years ago, because the processes of the agencies showed considerable similarities. However, these similarities turned out to be far fewer than expected and every local agency had a list with arguments why its deviations had to stay. The result was IT systems which had more custom code than common code.

The SCC was in the meantime held accountable for the accompanying budget and time-overruns. Efforts to not make the same mistake for the next IT system, resulted in months of discussion with the local agencies about which functionalities to incorporate and which not. And if there was agreement on the functionalities they were at such a high detail level that designers and programmers could only guess the underlying requirement…. and thus build something that was rejected during functional testing.

The combination of semi-independence, high budgets and local processes which were developed over several hundreds of years, resulted in a failure of transforming the intention to cooperate into a success. This combined with the ‘client-supplier’ relationship which came with the establishment of the SSC, resulted in local agencies putting all the blame with the SSC. The SSC got paid after all to build and operate IT services, and the client is always right….. (illustration 1)

The employees of the SSC had by now the feeling to be stuck between hammer and anvil.

Getting the Principal and SSC closer together

The cooperation between the local agencies and the SSC is depicted in illustration 2. Together, the yearly long term targets are defined, which the SSC is then expected to execute, with the help of the local agencies.

The first solution direction is depicted as option 1 in illustration 3. In other words, the various activities related to governance, information management and functional management are centralized in a ‘director-function’ between local agencies (the principles) and the SSC (the ‘supplier’). This ensures that the fights among the local agencies regarding their differentiations in demand are not fought in the front garden of the SSC.

The SSC is able to focus on building and operating IT services. Building functionalities which have been streamlined and made uniform before communicating them to the SSC. The ‘director function’ was made responsible for alignment of demand-supply and reported directly to the overall government board of the local agencies. Escalations were thus dealt with within the business domain and did not reach the SSC.

Illustration 3, option 2 provides another solution to reduce the tension between SSC and the principles. The second option focuses on a (temporarily) cut of the ambition level by cutting the IT portfolio in two parts. One part is build and operated by the SSC while the remaining services are decentralized again. For this second part a local agency can source independently from the SSC and other local agencies. To apply this second approach for the area’s with the highest level of differentiation, the overall tension level would be reduced.

By focusing the energy to IT services where it is more likely to be successful, a positive feedback spiral would in time lead to a situation where also more difficult challenges could be addressed. In time the end result would be the same, it would just take more time. And time is not a short supply in most government agencies.

The additional cost which accompanying these two options will in practice disappear against the budget overruns on ongoing projects which are executed in the ‘old’ fashion. The impact will therefore be mostly on the organizational side. With option 1, the SCC will have to split off its capabilities regarding information management and functional management as it would become part of the new director function in the business domain. With option 2, the shared service center would lose (temporarily) part of its portfolio and this resources and budget.

In the third part I will write about the application of both approaches and provide some general guidelines based on my experience.