At heart, the Strategy Unit has a very simple belief. We believe that the application of critical thinking and evidence improves services – and that this improves outcomes. At a headline level, this is our account of the way we add value. It is our theory of change.

Every service, every innovation, every policy, every organisation (every individual!) should have a similar account. Yet clarity about what is to be achieved, why and how is all too often:

  • Entirely missing. Perhaps the people involved can describe what they do, but not why they do it. The result typical is confusion and poor implementation; or
  • Partial and fudged. There may be multiple beliefs; these will vary in legitimacy and congruence. Sometimes stakeholders with very different aims will believe that the initiative is ‘delivering for them’. The result here is usually inconclusive: no-one knows what was being tested and so cannot make any assessment of success.

Each of us could think of a policy or programme, current or historic, that matches these descriptions.

To avoid this, the Strategy Unit advocates strongly for investing time and thought at the design stage. Getting thinking straight before moving into action – including knowing where uncertainties, assumptions and risks lie – is vastly more cost-effective than proceeding without going through these disciplines. This is not to say that every detail must be worked out – many initiatives require agility and course-correction – but a basic theory of change should be a pre-requisite. 

Logic models are a useful tool for setting this theory out. They are based on a simple set of questions, such as: 

  • What problem(s) are we attempting to address?
  • Given this, what difference are we trying to make? What outcomes do we want to see?
  • Given this, what should we do? And why (by what mechanism) do we think these activities will achieve our outcomes?
  • What resources do we need to implement our activities?

[Fraser explains in this short video how the set of questions can be defined]

One very quick way to save money would be to prevent the launch of any initiative that could not – as a minimum - provide a logically coherent set of answers.