David Ainsworth's article on the impact of impact measurement is good and timely. It is an important challenge to those of us who work in this space, but also to those who claim to not see any value in the impact question at all.

At Renaisi we're beginning to see both a change in the work we're being asked to do, but I also think that we're beginning to change how we think about the problem ourselves.

We are running an event on Thursday this week to discuss this question with some of our peers and colleagues from across the sector. I will write more after the event about what comes out of it, but I want to introduce three issues that I think are problematic, and which fit neatly with David's article:

  1. The incentives to ask this question are problematic. Impact questions are almost only asked when a funder appears on the scene, and as I believe and David stressed - measuring impact does not lead to money.
  2. The system that has evolved to 'service' this question, both internally within organisations and also external to them (consultants like me) are all complicit in not answering the questions that David poses. Does doing this lead to better outcomes, and if not, are we really asking the right questions?
  3. Finally, I think that the intellectual work that has shaped the impact agenda all sits within the 'evidence-quality' space, and that has negatively skewed the debate. Evidence quality is well grounded in social research principles but does not neatly transfer over to service improvement and the day to day working of charities. I don't mean that the quality of your evidence doesn't matter, but starting from that point means that you don't engage in five or six other really important questions. Evidence quality should be one, not the sole, pillar of your impact measurement approach. David poses some of the other questions in his article.

I'm keen, and believe that plenty of others are keen, to explore these challenges and try and think what to do to improve this situation. 

Impact questions matter greatly, but going about it in the wrong way is potentially more destructive to the debate as a whole, than not doing it at all. And I think not doing it at all is just not an option.