Ask any team under pressure to innovate and efficiency is often the first thing reached for: fewer steps, lower cost, less duplication. It is an understandable instinct, and in the public sector the pressure to find it is real. But efficiency is a measure, not a mission. It tells you how well you are doing something, not whether you should be doing it at all.
Take our work with the Ministry of Justice. Across a sentence, prison staff ask people in prison the same questions at many different points in time: how they are settling in, how they are coping, what they hope to do for work once they are released. Through an efficiency lens this is obvious waste. Ask once, record the answer, stop repeating yourself.
It is not waste. People in prison are not static, and neither are their answers. Asked early in a sentence or within earshot of others, someone might give a thin, guarded reply. Asked again months later, by a member of staff they have come to trust, the answer can be completely different.
Repetition is how staff notice that change. It builds trust. It puts a spotlight on things like wellbeing, so they start to register as something that matters. And it surfaces the inconsistencies that flag risk: a release address that sits a little too close to a victim, a story that quietly stops adding up.
So we were fierce about protecting what that repetition was doing. We were just as conscious that frontline staff are stretched, and that 'ask everything, every time' is its own kind of failure. The work was not to strip the repetition out, nor to pile more of it on. It was to separate repetition that earns its place from duplication born from poor coordination.
The same tension runs through related work connecting data to speed up benefit and job applications, so that people leaving prison on probation receive money they badly need far sooner. The opportunity there is real and significant. But resettlement is built on regaining agency, confidence and resilience. These considerations have to be balanced and weighed up: the streamlining and backend automations against what is lost if the person is designed out of their own application.
That is the danger in treating innovation as the blind pursuit of efficiencies. Efficiency optimises what is already there. It makes a process faster and cheaper without ever asking whether the process is right, or what the apparently wasteful parts are quietly holding together.
Real innovation starts a step earlier, with a harder question: what is this for, does it work, and what less obvious role might it be playing? It helps to treat efficiency as an intervention like any other, not a safe default.
The blind pursuit of micro-efficiencies can introduce unintended consequences as readily as anything else, only faster.