Everything in part one was produced without a day of consultant analysis. We could quote a twenty-six-week programme off it today. It would be priced against eleven open questions that could move a rating. But the assessment already lists every one of them, ranked, with the rating-movers flagged — so settling the reading is not an investigation to be scoped. It is a working session with the people who know, and the platform recalculates everything from the answers.
One and a half consultant days, inside one week, at £1,725. The recalculated assessment generates a roadmap with timescales, which we bring back to work through and agree with you — next week, not next quarter.
The case for change
The August assessment read every capability twice — once for the artefacts a mature pipeline requires, once for whether people still hit the problems those artefacts exist to prevent. Behaviour holds wherever something checks it, and drifts or stops wherever nothing does.
Level by level, across all seven capabilities. Top bar: the state of the artefacts that level requires. Bottom bar: how many of that level’s problems people report living with.
Not one Level 4 artefact was confirmed fully in place, and nine in ten of the Level 4 problems are being reported right now. Nothing here was abandoned. It was started, checked at the gate, and never watched again.
Only five items across all 73 have an artefact confirmed fully in place and the problem still reported. The two robust ones are both in Pipeline Foundations and both are between-gate: practice followed at the gates and worked around in between (0.79), records complete at transition and stale in between (0.75).
Demand and Pipeline Creation is the mirror image. Not one Level 2 or Level 3 problem reported — no friction, no complaint — while source recording, the coverage plan and in-period tracking all sit half-built. Nothing signalled it was unfinished. It scores 1.7 and sets the firm-wide level.
Twenty-three artefacts are absent and nineteen of them are feedback loops. Sorted by what you would buy, twelve of the twenty recommendations build the observation layer and eight finish an artefact. The eight are what the floor rests on. The twelve are what stops them going stale again.
The next step
AXAT carries every unsettled question forward, ranked, with the eleven that would move a rating flagged. Nothing needs scoping and there is no write-up phase — the work is defined before day one, and each answer is recorded in the session it is settled in.
| # | Step | What happens | Time |
|---|---|---|---|
| 01 | Review what is still open | Forty-two questions, already listed and ranked, with the eleven rating-movers flagged. Twenty-seven were never checked; fifteen are disagreements between respondents. | ~30 mins |
| 02 | Work the list with stakeholders | Evidence sessions against the named artefacts — mostly “show me” rather than “tell me”, because most of the open questions are about how much of a half-built artefact actually exists. The two people who did not respond come in here, and so does the CRM contradiction on page 6: a manager opens the system and shows us. | ~1 day |
| 03 | Record the outcomes in TheAX | In the session as each question is settled, not written up afterwards. Every answer carries its source, so the corrected reading is auditable rather than asserted. | ~1 hour |
| 04 | The platform recalculates | Corrected scores, corrected floor, corrected classification of every root cause, regenerated outputs. No re-analysis and no second report to write. | ~10 mins |
| 05 | The roadmap | Sequence and timescales generated from the recalculated assessment rather than from a separate planning exercise. Brought back as a draft to work through and agree with you, not a plan to sign off. | ~2 hours |
| One and a half consultant days inside one week, at £1,150 per day. Forty-two questions over about six hours of session time is roughly one every eight minutes. | £1,725 | ||
A corrected assessment at materially higher confidence, every settled answer carrying its source, and a roadmap with timescales generated from it and agreed with you. All of it is yours whether or not you commission anything further — including the roadmap, which you are free to put out to tender.
That the problem is larger than it looks — three capabilities carry confidence below 85% and could drop. Or smaller: the 29 half-built artefacts may be closer to finished than reported. Both outcomes are worth £1,725 and a week to know before committing six figures.
Indicative
What the settled reading is likely to require, based on this first pass. Steps two to four are indicative — the recalculated assessment generates the sequence and the timescales, and those are what the programme gets priced against.
| Step | Activity | Timing | Effort | Fee | Basis |
|---|---|---|---|---|---|
| One | Settle the reading42 questions worked through, recalculated by the platform, roadmap agreed | week 1 | 1.5 days | £1,725 | Committed |
| Two | Install the observation layerSampling between gates, cadence held, ownership named, loops closed — the twelve | weeks 2–13 | 30–38 days | £34,500–£43,700 | From the roadmap |
| Three | Finish and build the eight artefactsInside that layer, not after it | weeks 4–15 | 44–54 days | £50,600–£62,100 | From the roadmap |
| Four | Remeasure on the same instrument | weeks 16–17 | 4 days | £4,600 | From the roadmap |
| Indicative total across all four steps | 79.5–97.5 days | £91,425–£112,125 | — | ||
This organisation builds things and then loses them — that is the finding in part one, stated as a sequence. Sampling, cadence and ownership go in alongside the first artefacts, not after the last one. Build first and step three quietly recreates the problem step one diagnosed.
Quote a fixed price for steps two to four today. Any firm that does is pricing against a reading it has not checked — and here that reading is a week and a day and a half away, so there is no excuse for guessing. The range above is wide because eleven rating-movers are open and 29 artefacts are half-built by an unknown amount.
All 42 open questions closed, or explicitly carried with a stated reason. Outcomes recorded in TheAX with their source, the assessment recalculated, and a roadmap generated from it and worked through with you. £1,725 fixed: if it takes two days, that is our problem.
A named floor target, a set of countable adoption thresholds measured weekly from your CRM, and a remeasurement on the same instrument. We will put fee at risk against the artefacts and commit jointly on the behaviours — one week from now, against a reading we have checked.
Level 4. Nineteen of the 21 Level 4 problems are being reported right now and no Level 4 artefact was found complete. That is a second programme, and the first honest conversation about it is a year out.
A commitment made against an unchecked reading is not a commitment. It is a guess with a signature on it. One week from now we will give you one worth holding us to.
Conditions
Your baseline was attributed, not anonymous. Six named people told us how their pipeline actually runs. There is an obvious temptation, later, to use those answers as evidence about the people who gave them.
Assessment responses are never an input to any decision about an individual. If they were, the remeasurement would be worthless — everyone would answer defensively — and you would have punished candour you asked for. Any judgement about people rests on observed practice against a written standard, and that standard does not exist yet: the capability framework is one of the eight artefacts still to be built.
| The two stakeholders who did not respondOne session each | Without them, Technology and Data stays at 73% confidence and the Demand and Pipeline Creation floor stays unproven — and that floor is what sets the firm-wide level. |
| Diaries cleared enough to do it in five daysThe real constraint | A day and a half only works if the sessions can be scheduled back to back. Spread them over a month and the elapsed time, not the effort, becomes the cost. |
| A sales manager who can open the CRMIn the session, not afterwards | The CRM contradiction cannot be settled by asking the same question again. It is settled by someone showing us the system and saying whether trackers exist alongside it. |
| Whoever can say what was actually built, and whenOften not the sponsor | Twenty-nine artefacts are half-built and most of the 27 never-checked questions are about them. Someone knows why each one stopped, and that is the fastest route to settling it. |
Everything in part one rests on one claim: that behaviour holds where something checks it and drifts where nothing does. If the settle week confirms Rob S — if half the team really does run trackers outside the CRM — this is partly a tooling-fit and compliance problem, not purely an observation one.
What we do: We have named the falsifier rather than burying it. It is one of the fifteen disagreements and it is settled inside the stakeholder sessions, by asking someone to open the system rather than by asking the question again. If it goes against us, the step two scope changes and we will say so when the roadmap is generated.
Validating evidence usually finds less than was reported. Only 30% has been checked, and three capabilities carry confidence below 85%. Technology and Data at 3.3 is the most exposed — it is the highest score on the fewest responses.
What we do: Better to learn that for £1,725 in week one than in month four of a fixed-price programme. This is the central argument for the settle week existing at all, and we would rather present it as a risk than as a selling point.
A corrected assessment with no delivery behind it is a report. Sunshine Power already has plenty of things that were produced, filed and never carried through — twenty-nine of them are half-built right now, and they are listed in A4.
What we do: The roadmap is generated with sequence and timescales attached and worked through with you in the same week, so it arrives agreed rather than filed — and it can be acted on by us, by you, or by whoever you put it out to.
The deepest risk, and the one the assessment itself demonstrates. Sampling, cadence and named ownership are exactly the kind of thing this organisation starts and stops.
What we do: Stage two is sequenced ahead of stage three deliberately, and every observation item is signed off on evidenced use rather than on delivery — a sample actually taken, a review actually held with attendance recorded, a decision actually logged.