Method

One instrument, from before the LOI to twelve months after close.

What follows is what the instrument measures and how a value gets made. The last section is the one worth reading twice.

What is measured

Four pillars, each drawn from a distinct thread of the M&A integration literature. They are not four names for the same thing: a deal can be clean on one and breaking on another, and which one breaks changes what you do about it.

A · Cultural distance — Teerikangas · Rouzies · Angwin
B · Leadership retention — Durand · Krug & Hegarty
C · Employee sense-making — Kroon · Schweiger & DeNisi
D · Comms coherence — Larsson & Lubatkin

How it is measured

A fixed battery. The same questions, in the same wording, on the same scales, on every deal — that is what makes a pre-LOI reading and a twelve-month reading comparable at all.

Three item types: a 0–10 score, an open text prompt, and a short multiple choice. Score items carry the scale in the question itself rather than in a legend, so a respondent answering on a phone is answering the same question as one answering at a desk.

Audiences differ by stage. A leadership round-table is six to eight people; an organisational survey is everyone. The instrument does not change between them — the sample does.

Why we show the spread instead of one number

Six leaders answering 3, 3, 4, 8, 9, 9 average to 6. The average is the least useful thing in that sentence. What it hides is that the leadership team does not agree about its own organisation — and for a deal team, disagreement is the more actionable finding, because it names where the conversation has not happened yet.

So every value carries its distribution and the number of responses behind it. Not as a footnote: the count sits with the value, because a value read without its n is a value read wrong.

We do not put a confidence interval around it either. On eight to forty respondents an interval would be its own kind of false precision — a decimal-place answer to a question the sample cannot resolve. Showing the shape of the responses is honest at that size; a ± is not.

Below a minimum cohort the shape is withheld entirely, and you get the count and the middle value instead. A distribution over a handful of people is a fingerprint: in a team of eight, one answer sitting alone at the bottom identifies whoever gave it. Protecting the respondent is not a caveat on the method here — it is the reason anyone answers honestly.

How a friction point is derived

A friction point starts as responses, not as an opinion. It attaches to one assessment and one pillar, carries a severity, an owner and the evidence it was drawn from, and can be opened back to the answers underneath it at any point.

That last property is the one that matters in a deal committee. A finding nobody can trace is a finding somebody will dispute, and disputing it costs more meeting time than gathering it did.

How pre-LOI assessment works without access to employees

Before an LOI there is no contact with the target's people, and no survey. What exists is public and observable signal, scored on the same instrument as every later stage so that the first reading and the last are on one axis.

It is a thinner reading than a survey and it is presented as one. The value of taking it is not that it is complete — it is that when diligence opens and the survey runs, you have something to compare against that predates your own involvement.

The research it rests on

Every pillar attribution above resolves to a full citation — authors, year, publication and a link — on the research page. If a name appears on a Cuddly surface without a paper behind it, that is a defect, and a test fails the build for it.

What this method does not do

It does not predict outcomes. We have no completion data, so we make no claim that a given gap causes a given delay.

It does not benchmark you against other companies' deals. That dataset does not exist, and we will not market it as coming soon.

It does not replace a licensed academic instrument, and it is not claimed to be better than one. The difference is that it runs on the deal clock and stays comparable across deals.

It does not produce a single culture score. A number that compresses an organisation into one figure is not something a deal committee can act on.