For combined authorities, agencies and growth hubs
"We put 200 companies through the programme" won’t survive the next settlement.
Throughput was acceptable evidence while budgets were growing. It won't be while they're being reallocated, and the programmes that can show movement will be the ones that keep their funding.
The commissioning problem, drawn
Not two hundred problems. One problem, forty six times.
A throughput figure tells you the money was spent. It cannot tell you what to spend it on next, because it contains no information about what is actually wrong.
Measure the same cohort on one instrument and the picture concentrates. The weakness is rarely spread evenly. It clusters, and the cluster is commissionable.
That is the difference between reporting a programme and designing the next one.
An illustration of how weakness concentrates in a cohort, not observed data.
What the report contains
The quarterly report, page by page.
Cohort movement
Every company measured at intake and exit. The proportion that moved, the average movement, and the stage where it happened.
Position and cause
Where each company stands, the named root cause, the intervention routed, and what changed at the next measurement.
One language
Every programme on the same published instrument, so findings from one partner mean the same as findings from another, and provision can be planned against what companies actually need.
The sentence you take to the next settlement changes. From "we put 200 companies through" to "we measured 200 companies, this proportion improved, and here is the intervention that did it".
Two ways to fund it
One of them needs no new budget line.
- The authority licenses it centrally and requires its programmes to use it.
- Or each delivery partner funds it from support money already allocated to them, at no new cost to the authority.
Because the method is published, a delivery partner can also run it themselves. That’s deliberate: adoption matters more to us than seat count, and a partner running the published instrument produces data as comparable as one we run.
What every company in your cohort receives
A ten-page read your delivery partners can act on: the position, the named root cause, the category of support indicated, and a plan with a 90 day proof point.
Free to them, whether or not they ever speak to you again. The same result is what aggregates into your reporting, so there is one artefact and two audiences rather than a company view and a separate management view that can disagree.
What changes in the job
- Provision is commissioned against evidence. If one root cause dominates the portfolio, you can fund against it
- Referrals stop being guesswork. The no wrong door promise gets a mechanism behind it
- Delivery partners speak one language. A finding from one means the same as a finding from another
- The impact report writes itself. Quarterly, from delivery, instead of commissioned every few years
- The outcomes framework gets a business support measure that is not throughput and not three years late
The payoff
| Measure | Today | After one cycle |
|---|---|---|
| Outcome evidence per programme | Throughput, self-reported by the provider | Movement per company, same instrument across providers |
| Findings across delivery partners | Different metrics, not combinable | One scale, so the portfolio can be read as a whole |
| Where support money should concentrate | Committee judgement | The stages and causes that recur across the portfolio, counted |
| The impact report itself | Commissioned, occasional, expensive | Produced quarterly as a by-product of delivery |
On assurance: the method, arithmetic and conformance vectors are public, which answers most of an algorithmic transparency or data protection review before it starts. It’s a measurement instrument, not an automated decision about any person or business.