Insights
Fixing the Organizational Idea Intake Gap
A state agency runs an employee suggestion programme. In its first eighteen months it received 132 ideas from a workforce of tens of thousands. Twelve were adopted. Those twelve produced roughly $7.2 million in verified savings in a single fiscal year. The savings figure gets quoted in the annual report. The interesting number is 132 — small not relative to the savings, but relative to the number of people who spend every working day watching a public service function badly and knowing precisely why.
That gap is not a motivation gap
There is good evidence about where ideas die. A field experiment across every ministry of the Ghanaian Civil Service found strongly hierarchical cultures in which junior staff felt unable to raise new ideas at all. Researchers trained bureaucrats to break problems into small, actionable solutions and to raise them with colleagues. Six to eighteen months later, the individually trained staff were measurably more likely to raise and discuss ideas, and administrative processes had improved. The part I find most instructive is what did not work. When the same training was delivered at division level, it largely failed. Divisions absorbed it into their existing hierarchy and produced plans that were grand, resource-hungry and Civil-Service-wide — the sort of proposal that requires someone else’s budget and therefore never happens. The individual version produced small, local, implementable changes. The collective version produced ambition with no owner.
Hierarchy did not reject the intervention
The constraint has moved. When the UK’s Department for Work and Pensions evaluated a generative AI assistant across 3,549 staff, it measured an average saving of nineteen minutes per person per day. Multiply that across a large agency and you have an enormous quantity of newly available human attention inside institutions whose formal channels can absorb perhaps a hundred ideas a year. We are increasing the supply of thinking time while leaving the intake capacity for thoughts exactly where it was. An idea has to survive a long journey: recognised as an idea by the person who has it, spoken aloud without professional risk, understood by a manager whose incentives it may complicate, translated into institutional language, then owned by someone with a budget. Most die at step one or two, silently, leaving no record. The intake number does not measure how many ideas exist. It measures how many survived the filter. If that is right, the design question for the next decade is not how to make people more creative. It is how to shorten that journey.
Three things follow, none expensive
Route ideas to the smallest unit that can act, not to the highest one that can approve — the Ghana result suggests scale is where ambition goes to die. Judge a suggestion system by the proportion of submissions that receive a reasoned answer, not by the proportion adopted; the first number is what people actually respond to. And accept partial thoughts. Most institutional intake forms demand a fully formed proposal with a cost-benefit case, which is a request for a finished product from someone who has noticed a problem. A quieter argument sits underneath all this. A suggestion programme is not an efficiency tool. It is a statement about whose observations the institution considers real. When a caseworker or a night-shift technician submits something and receives a serious reply, the organisation has said: your reading of this place counts as knowledge. That is a matter of dignity before it is a matter of savings — though, as those twelve ideas suggest, the savings follow. Every large organisation already contains the analysis it is currently paying consultants to produce. The difficulty was never generating it. It was building something capable of hearing it.