The missing variable: a reputation is measured by what it makes people do

This is the thesis of this cahier. It fits in a single sentence, and everything else follows from it: a reputation is worth only the behaviours it triggers. The rest — esteem, image, sentiment — is not false. It is simply what precedes the only thing that counts: the act.

Most executives steer their reputation the way one steers an opinion: they observe it, poll it, take heart or take fright. But an opinion is an input, not an outcome. The outcome is what the opinion makes the market do — and that, and that alone, has an economic effect.

The outcome is what the opinion makes the market do — and that, and that alone, has an economic effect.

Supportive behaviours

A market that holds you in good reputation does not merely think so. At one moment or another, it acts on that judgment, and that act has a value. Four behaviours capture the essence of that value for a company your size:

• Recommend. People talk about you in your absence, they refer you, they put you on the table when the subject comes up. This is reputation selling in your place.
• Apply. People choose you as an employer, they apply unprompted, they prefer your offer to higher pay elsewhere. This is reputation recruiting in your place.
• Grant the benefit of the doubt. When an incident occurs, people suspend judgment, give you time, do not switch at the first alarm. This is reputation protecting you.
• Pay, and stay. People accept your price when a competitor is cheaper, they renew without putting it back out to tender, they resist the pull of the market. This is reputation holding your margin.

These four behaviours share one decisive property: they are observable. You cannot directly measure the esteem a company inspires. You can measure its share of recommendation, its ratio of unsolicited applications, its resilience in a crisis, its renewal rate without a tender. Reputation stops being a state of mind; it becomes a set of accountable signals.

These four behaviours do not carry equal weight. Three carry the bulk of the value for an established company — recommend, apply, protect — and chapters 4, 5 and 6 take them one by one. The fourth, pay and stay, will have no chapter of its own: it reads through the other three as much as through your margins, and chapter 8 recovers it as a measurement signal rather than as a separate subject. This cahier therefore develops not four behaviour chapters, but the three that weigh most heavily on a profit-and-loss account.

Why the gap exists — and why it is costly

Between opinion and act, there is a gap. That gap is the real blind spot. A company can have an excellent opinion and a low propensity to behaviour: it is esteemed, but not referred; respected, but people apply elsewhere; called serious, but dropped at the first incident. This is the case of the executive in chapter 1 — and it is the most common case among companies that have tended their image without ever organising the acts that should have followed from it.

The reverse also exists, rarer and more precious: a company with a modest image but a very high propensity to behaviour. Unspectacular, but highly recommended. Discreet, but not one that people leave. Such companies have understood, often without putting it into words, that the useful reputation is not the one people admire — it is the one people act on.

Armor-Lux, or the reputation that gets contracts signed

Take the through-line of this collection. Armor-Lux, a Breton textile manufacturer, did not win its major workwear contracts — Brit Air, La Poste, the SNCF, Carrefour — because it was esteemed. It is esteemed, but esteem does not sign a contract. It won them because its reputation had come to do something: it had deposited, in one buyer after another, a settled conclusion — "French manufacturer, serious, dependable over time" — robust enough for a purchasing director to recommend it to their committee without having to re-justify it.

That is what a reputation at work is: not a good image, but a decision shortcut installed in the minds of those who decide. The buyer who renews an Armor-Lux contract does not redo the investigation. They act on a stock of conclusions they no longer need to verify — and it is that stock, not esteem, that has value.

The chronology of these contracts, consolidated in Cahier I, says exactly this: a reputation that, decade after decade, got different buyers to sign, in different markets, without the company having to reintroduce itself each time. This is the yield of a stock, not the effect of a campaign.

What this shift changes

Moving from "what do people think of us?" to "what does that judgment make people do?" is not a nuance. It is a change of instrument. The first question calls for an image barometer; the second calls for a measurement of behaviours. The first cheers or worries the committee; the second hands it a judgment call. For if reputation is measured by what it makes people do, then one can identify which of these behaviours are missing, and concentrate effort where the gap between opinion and act is most costly.

A reputation is not something to contemplate. It reads in the acts it produces — or in those it fails to produce, even though the opinion was good.

What remains is to take these behaviours one by one, starting with the one whose yield is highest and whose management is rarest: the one that sells in your place.