Methodology
Every question, every weight, and the band it lands in
Ten questions, ninety points, four possible results. This page prints all of it — including which answers are worth nothing — so the score can be argued with rather than believed.
The result is not gated. There is no form between the tenth question and the answer, and there never needs to be one for you to read this page: everything the instrument does is written below, so you can score yourself on paper and never touch the tool at all. That is deliberate. An assessment issued by somebody who sells the software has to be checkable or it is advertising with a dial on it.
The instrument
The ten questions, and what each is worth
Every question offers four answers. The healthiest answer earns the maximum shown; the worst earns zero in all ten.
| Question | What it actually measures | Max |
|---|---|---|
| How many different answers you would get if you asked four people how a purchase gets approved | Whether one process exists, or several share a name | 8 |
| How much of your customer list you would load into a new system unchanged | Whether migration is a task or a workstream | 10 |
| Who is accountable for the customer master — the record set, not the department | Whether ownership can refuse a duplicate; a department cannot | 10 |
| How much of the sponsor's week is in the diary, for the whole project | Whether decisions have a route, measured in calendar rather than enthusiasm | 12 |
| What the approved budget actually covers | Whether year two onwards exists in the business case | 8 |
| Where your requirements came from | Whether they can disqualify a vendor, which is the only useful thing they do | 10 |
| Whether the company has tried this before, and whether anyone wrote down why it stalled | Scar tissue, and whether it was ever examined | 6 |
| What happens when the software does something differently from the way you do it now | Whether the project is buying a system or commissioning your current process in licensed form | 10 |
| Who decides when two departments want incompatible things, and whether it holds | Whether a disagreement has an end | 8 |
| Who owns the system six months after go-live | Whether anyone will be left to measure what the business case promised | 8 |
The arithmetic
Adding it up
There is nothing clever here, and that is the point: the judgement in this instrument is entirely in the weights above and the thresholds below, not in the sum. Anything that looked like statistics in the middle would be decoration.
The bands are declared against a percentage rather than a raw score so that a question can be added later without rewriting every threshold. The number on the dial and the band beneath it are computed from the same percentage, so they cannot disagree.
From ten answers to one band
- score = the points on each of the ten answers you chose, added up
- the maximum available is 8 + 10 + 10 + 12 + 8 + 10 + 6 + 10 + 8 + 8 = 90
- percent = round(score ÷ 90 × 100)
- band = the highest threshold at or below that percentage
The weighting
Why the questions are not worth the same
The spread is narrow — six to twelve — but it is not arbitrary, and it encodes one argument you are entitled to disagree with.
Weighted heaviest
- Sponsor's diary time, at 12 — the single heaviest question. An ERP project generates several hundred small irreversible decisions, and a decision that waits four weeks is indistinguishable, from the outside, from a project running late.
- Data quality, master data ownership, where requirements came from, and what happens when the software disagrees with you — 10 each. Each of these is settled before a vendor is contacted and is expensive to revisit once one is on site charging by the day.
Weighted lightest
- Whether the company has tried before, at 6 — the lightest question, and the one most people expect to be heaviest.
- It is light because a first attempt is not a weakness: that answer scores 5 of the 6 available, barely below the company that failed once and wrote down why. What is actually being measured is whether the organisation examines itself, not whether it has scars. Failing repeatedly with nothing written down is what scores zero.
The results
The four bands, and where each one starts
Matched from the highest threshold downwards, so a score of 80 returns the top band and a score of 79 does not.
| Result | From | What it returns |
|---|---|---|
| Ready to start | 80–100 | Nothing here says wait. Spend the advantage on requirements written from observed work, demos scripted on your own data, and the contract terms that matter while you still have leverage. |
| Ready, with two or three things to close first | 60–79 | The foundation holds. The flagged items are work rather than caveats — each is cheap now and expensive once a vendor is on site. |
| Not yet | 38–59 | The gaps are real and none of them need software. Name an owner per master data set, agree who settles a cross-department dispute, walk one real transaction per process. |
| Do not start yet | 0–37 | Buying now would make the current mess permanent, licensed and annual. The disagreements would move into one database and acquire the authority of a system. |
Where it stops
What this score does not know
The bounds matter more than the number. Every one of these is a real limitation, not a disclaimer.
- It is entirely self-reported, and it has no way to detect a company answering as it wishes it were. The instructions ask for the process as it actually runs — including approvals given in a corridor — because a flattering answer produces a flattering score and no information.
- It says nothing about which system to buy. Every question here would matter identically whichever product was eventually chosen, which is what makes it vendor-neutral in a way a feature comparison cannot be.
- It does not know your industry, your size, your country or your transaction volume. A ten-question instrument that claimed to weigh those would be pretending to a precision it has not got.
- The number is the least useful thing it produces. Every answer below the maximum attaches a specific flag naming what that answer costs and what to do about it, and those flags are the actual output — two companies on the same score can have entirely different work in front of them.
- It measures readiness to start, not likelihood of success. A company can score well and still run a bad project; nothing here can see how the organisation behaves under pressure eight months in.
If the flags look familiar
The flags are the part worth a conversation
If your score came back with three or four flags and you want a second opinion on which to close first, that is a short conversation and it does not require buying anything.
Monday to Friday, 9:00 AM – 6:00 PM (GST)
Prefer we call you?
Leave your WhatsApp number and we will reach out.