Will shared services ever work?

Last week the UK Government launched a new strategy for shared business services (common things such as finance and human resources). The general reaction seemed to be quite negative, at least in my filter bubble. Most of the criticisms are probably reasonable but that may be because there are so few examples of sharing successfully at this scale in other countries or sectors.

I am hoping to join a discussion about all this at UKGovCamp at the weekend and some in Government are keen to hear what campers think.

Now that would be really helpful. I can't go but would be really grateful if you could raise this.

— Liam Maxwell (@liammax) January 14, 2018

In the meantime here are some potential challenges and responses to consider. Most of these are relevant for sharing business services in any sector.

Economies of scale have limits

I have not seen the cost benefit analysis for this new strategy, if there is one, but I have worked on a few for shared business services and reviewed many more. Most have an implicit assumption that more sharing always produces benefits. Most cases recognise diminishing returns but it is very rare to consider negative returns (where marginal increases in cost exceed the marginal benefits). This must be a practical consideration at the scale of UK government.

The thresholds vary for different types of service (e.g. I have written about this for IT in the past) but, as a rough guide, pooling services for organisations with a combined workforce of 10,000 - 25,000 people can generally be made to work and deliver an attractive benefits profile. At a smaller scale there will almost certainly be more valuable investment opportunities in your portfolio. At a larger scale the barriers get greater and greater and the net benefits may quickly evaporate.

Some large multinationals have rowed back from sharing business services across their whole organisation due to these issues. Some individual Departments in the UK are already at this scale even if they don’t share services with others.

Shared ERP software has very little value

I used to work for one of the big ERP vendors so I am one of the few people who have used a raw installation of one of these suites of software. Many people will be surprised that you can do almost nothing with the system at first. These suites only start to work when you feed them vast amounts of information including details about your customers, staff, suppliers, assets, accounting practices, roles, operating structure, business policies and working practices.This is a kind of digital twin of the operations your service will support.

The technology costs are usually dwarfed by the upfront costs of collecting, translating and testing this information and the on-going costs of maintaining it. The more ways of working a shared service needs to support the more information, and therefore cost, will be consumed. Adding a small organisation with distinct processes to a larger shared service can destroy the business case for both.

Work with, not against, the natural limits to shared business services

In theory, it would be possible to mobilise a programme that could deliver cross-Government shared services but you would need to pick a truly exception team of people from a number of different Departments and suppliers. These people do exist but, unfortunately, they are already running operations and change programmes with much greater impact than back-office business services.

A more practical response would be to focus on capturing the large share of benefits available at a smaller scale. This would mean:

• Identifying several clusters of organisations at the 10,000 - 25,000 staff scale.
• Appoint a business services director for each cluster.
• Transfer the business service operations and funding to the director.
• Set ambitious goals for improvements but allow the directors to take different approaches for their cluster.

This might not release all of the theoretical benefits of more substantial sharing but is at a scale that the teams involved can manage. It will take less than half of the 10 year strategy timetable offering the opportunity to go further if the evidence and capability is there. It will also allow alternative approaches to be compared in practice including using different technology and delivery methods.

Watch this space

The next blog post provides some feebback on the discussions at UKGovCamp.