About
Where the numbers come from
A business performance framework that helps owners and consultants identify what deserves attention next: a structured self-assessment, research-informed opportunity benchmarks, and a plan that connects each priority to human action, consultant support or automation.
Every score on this site is taken from one research document, version 1.0, dated 22 August 2026. Nothing is calculated here that the document did not state, and nothing is rounded differently from the way the document rounds it.
Framework v1.0 · benchmark research dated 22 August 2026 · research-informed prioritisation, not a validated diagnostic
The three numbers
Time released
Typical potential to release owner or staff time, cut delay and rework, or add capacity if the area is materially improved.
Financial upside
Typical potential to improve revenue, margin, cash or cost efficiency — or avoid material loss — relative to the resources required.
Automation potential
How much of the recurring work, monitoring, analysis or execution can typically be automated or AI-augmented, with appropriate controls.
What these numbers are not
They are structured benchmark judgements about typical cross-business opportunity. They are not guaranteed returns, not forecasts, and never a particular business’s own result. What you actually get depends on your industry, size, current maturity, transaction volume, labour cost, margins, system readiness, risk and how well the work is done.
The research deliberately refuses to pre-score how critical any area is, on the grounds that criticality has to be assessed against an individual business and cannot be responsibly assigned in advance. That is the gap the audit fills: you supply the half about your business, and the ranking comes from putting the two together.
The arithmetic
A major category’s score is the sum of its sub-category scores. A general category’s score is the sum of its major categories. Across the whole framework that is 1847/2780 for time, 2229/2780 for financial and 2043/2780 for automation.
Averages appear alongside the sums only so categories containing different numbers of sub-categories can be compared fairly. They never replace the sum, and they are the document’s own stated averages rather than recalculated ones — where a recalculation would round differently, the research wins.
How the audit ranks an area
One line of arithmetic, stated so it can be argued with: an area’s rank is (10 − your rating) × the average of its three benchmark scores — time released, financial upside and automation potential, weighted equally. A weak area with little typical upside is not urgent; a strong area with high upside is already being harvested. The product of the two orders the list and is never shown as a score, because it is not one.
It is research-informed prioritisation, not a validated diagnostic. The benchmarks are cross-business judgements; your rating is your own honesty. Neither has been calibrated against outcomes yet, and the site says so wherever a ranking appears.
The written guidance
The explanations and automation advice attached to each of the 278 sub-categories are ours, not the research document’s. The document scores; it does not explain. The guidance is written for an owner rather than a consultant, and it is deliberately honest about the areas where automation does not help — culture, trust, succession and leadership continuity score low for automation potential for good reasons, and the advice says so rather than inventing tooling for problems that are not tooling problems.
Quoting and reuse
Quote and cite freely. Attribution to the 26-Point Checklist is appreciated but not required. Please do not present the benchmark scores as any individual business’s own result — they are researched cross-business averages, and a number lifted out of context becomes a claim the research never made. The written guidance is ours; the scores are the research’s. The machine-readable copies (ontology.jsonld, llms-full.txt) carry these terms with them.
Read the research
Everything above is a summary. If you want the scoring method in full, the calibration bands, and the 22-source evidence library the judgement calls were made against, it is all on one page — nothing here is asserted without a named source behind it.