We need to talk about spend controls

Dafydd Vaughan on 10 April 2017

Recently, the National Audit Office (NAO) released a report on the Digital Transformation of government.

In it, they devote a section to the approach and results of the government’s control of technology spend.

As always happens, the focus is on “savings”. After all, what’s the point of controlling spend if you don’t make savings?

People love savings, government in particular. The bigger the savings the better.

Focussing on savings can quickly lead you to the conclusion that you should spend most of your time and effort controlling the big projects – the ones where there’s lots of money, and potentially lots of savings.

Indeed, that’s exactly what the NAO’s report does. It highlights that GDS spent nearly half its time (47%) on projects that generate little savings. It criticizes GDS for this.

Unfortunately, the approach taken by the NAO auditors and others is incredibly naive.

While immediate savings are definitely a benefit of spend controls, there are many other benefits which don’t necessarily lead to direct savings. Some of these benefits are more difficult to quantify, but no less important, and may eventually lead to a better overall cost saving later on.

Connecting projects, teams and ideas together

In a large organisation, you’re very likely to find people doing the same, or similar things.

You might find different departments building related services, or replacing similar tech stacks or buying the same thing.

Introducing spend controls helps you to identify where this is happening and gives you the opportunity to break down those project and department silos.

You might be able to join projects together so they can reuse each other’s work or share the outcome – helping to reduce duplication and overall cost.

You can get the projects to share research or knowledge about approaches that work well – helping to make the work more efficient.

The effect of creating these connections is difficult to quantify, but can massively help projects deliver successfully.

Making course corrections

Projects don’t always start out as big expensive things. Picking them up early gives you an opportunity to tweak their course and plans.

You could direct them towards different procurement routes; or suggest different ways of undertaking the work; or split the work up into more logical steps.

You could stop projects that don’t have a genuine need before they get too far.

Intervening early, before projects grow, can help them from becoming unnecessarily big in the first place.

This is particularly true of agile projects – these shouldn’t be huge massive multi-million pound pieces of work.

They should be small, incremental pieces which can be tweaked and changed as they go along.

This reduces the likelihood of failure, but also means you can stop them before they spend too much money.

It’s why GDS approves agile work in phases – no one approval is too big.

Large projects need large interventions

Large IT projects aren’t easy to fix – it takes lots of effort to turn them around

It’s often simpler (and cheaper) to handle that intervention much earlier, before projects become big unstoppable supertankers.

The typical large IT programmes are usually significantly progressed in their thinking before getting caught by the controls process that getting them to consider alternative approaches or solutions takes lots of effort.

More often than not, you’ll find that large projects have already done most of the work to buy the solution they want, which means you’ve got to unpick that as well.

You need people with the right experience, skills and authority for a successful intervention. You need the right political backing.

Yes, you can make big savings, but you need to invest in the infrastructure to do it.

An alternative view

The NAO suggests that GDS spent it’s time badly – that they should have focussed on the other 53% of projects which contributed 99% of the savings pot.

They ignore the other benefits gained from supporting the smaller things; the early interventions; the course corrections; the shared knowledge and understanding from putting teams and projects together.

I’d like to offer an alternative view.

Spend controls and the Digital Service Standard have existed for over 5 years now. The rules aren’t new.

Despite this, GDS is still having to spend more than half of it’s time dealing with large £1million plus IT projects. And it’s still able to make more than £1.2 BILLION in savings from them.

After 5 years of effort, government is still addicted to big expensive IT projects.

What does that say about the technology maturity in government departments?

Any views stated here are my own and not those of my employer unless otherwise stated.