Almost every leadership team has a document like this somewhere.
It came out of an offsite, or an agency engagement, or a workshop with a facilitator who was genuinely good. Everyone was in the room. Everyone nodded. The document was circulated, and for about three weeks people referred to it.
Then it stopped. Not dramatically. Nobody rejected it. It simply stopped being the thing anyone reached for, and eighteen months later the leadership team describes the company four different ways again, and someone books another offsite.
The usual explanations are that the consultant was mediocre, or the company didn’t follow through, or the framework was wrong. Occasionally one of those is true. Usually none of them are, and the real reason is structural.
Two things get written down, and afterwards you cannot tell them apart
In most of this work, two very different kinds of claim end up in the same document, formatted identically.
Some claims were tested. Someone said the company was faster than the alternatives, was asked how a customer would know that without being told, and produced something — a number, a process, a comparison that held. That claim earned its place.
Other claims were not tested. They arrived because they sounded right, because the founder has always said them, or because nobody in the room wanted to be the person who challenged the CEO’s favorite sentence at four in the afternoon.
Both end up as bullet points in the same typeface.
Six months later the document meets what actually happens in a company. A campaign needs a headline. A salesperson needs an answer to a hard question. A partner pitch needs a differentiator. Someone reaches into the document and pulls out a line — and they have no way of knowing whether the line they picked was load-bearing or decorative.
So the tested claims get used as decoration, and the decorative ones get cited as though they were proven. Both fail. And the document quietly stops being trusted, because it produced a result that didn’t hold and nobody can reconstruct why.
That is not a discipline failure. It is what happens when evidence and preference are recorded in the same document with no marking to distinguish them.
The bundle is what makes it possible
This happens most reliably when the work is bundled.
When one engagement covers the value proposition and the naming and the identity and the tagline and the messaging, the pieces stop being separable. They are all in the same deck, delivered on the same day, by the same people. Some of that work has an evidence standard. Some of it is a matter of judgment — naming, identity, the feel of a thing — which is a genuinely harder discipline in the ways this one is easy, and which this method deliberately refuses to exercise. It is not lesser work. It is work with a different standard of proof.
Mixed together, the taste work borrows the authority of the tested work, and the tested work inherits the arguability of the taste work. Everything becomes equally negotiable, which means everything becomes equally ignorable.
There is a second, sharper failure in the bundle. If a leader has already privately decided something in one of those adjacent areas — the name is staying, the look is staying — that decision doesn’t stay in its lane. It reaches backwards and shapes every other answer, so that the value proposition ends up justifying a decision nobody in the room admitted to making. Frequently the person holding that pre-decision cannot see it in themselves.
The only reliable protection is to keep the adjacent work out entirely. Not to handle it carefully. To exclude it.
Which brings us to the part that is genuinely uncomfortable
If the work has to be kept narrow, someone has to enforce the narrowness — in the room, in real time, against people who outrank them and are paying them.
Consider who is usually in that chair.
An agency should not have to enforce it, and it is unreasonable to ask them to. The adjacent work is their next invoice. When a client hands the agency the value-proposition question, they are asking the agency to referee a match it is playing in — and then treating the result as soft when it comes back hedged. Most agencies know this perfectly well. The good ones say so out loud, and get told to run the workshop anyway.
An internal executive cannot enforce it. They have to work with these people on Monday. They report to someone in the room, or manage someone in the room, or want a job someone in the room controls. Asking the CEO to substantiate a claim they’ve been making publicly for six years is a career act, not a facilitation act.
The CEO cannot enforce it. They are the person whose claims most need testing, and they cannot both make the argument and rule on it.
A fractional executive is closer, and frequently spots the problem first — but if they are also delivering the marketing that follows, they are back inside the same conflict.
This is not a comment on anyone’s integrity. It is a statement about position. A referee cannot play in the match. Not because referees are more virtuous than players, but because the role only works when the person holding it has nothing riding on the score.
The tenet
The work must be run by someone with nothing else to sell you, and no stake in the outcome except that the answer is true.
That is the whole doctrine, and everything else follows from it. Single service, because a second service is a stake. No implementation, because implementation is a stake. No naming or branding attached, because those are stakes too, and because they contaminate. The person in the chair rules on claims and does not make them.
It also means the discipline has to be willing to be expensive relative to its size and narrow relative to its market, because both of those are what independence costs.
What the outside party does not do either
Independence cuts in the other direction too, and it leaves real work on the table for someone else.
The method specifies the preparation the room depends on: the interviews, the input gathering, the customer evidence, the competitive material a leadership team needs in front of it before the day starts. It does not specify who performs that work, and we do not perform it. In most engagements the natural party is the agency, the fractional executive, or the internal marketing lead who already knows the business and already has the relationships. The worksheets that structure that preparation are built to be handed over and used by them.
The same is true afterwards. The day produces a decision and the documentation of it. Turning that into campaigns, sites, collateral and sales enablement is somebody else’s job by design, and it is usually better done by the people who were going to do it anyway. A tested value proposition makes that work easier to scope, easier to defend internally, and much harder for a client to relitigate six months later.
So the boundary is not a claim that this is the important work and the rest is downstream. It is a division of labor with a specific reason: the party that rules on the claims cannot be the party that profits from what the claims are used for.
What this rules out, including for us
An honest doctrine has to say what it forbids its own practitioners.
It means we cannot sell you the campaign afterwards, which is where the money in this industry usually is. It means we turn down engagements where the people who must agree cannot all be in the room, because a consensus that needs ratifying later is not a consensus. It means a day can end with a claim marked unproven, and that is a real outcome rather than a failed session.
A disclosure, since the argument applies to me. I also run Mediathink, which does marketing work. That is precisely the conflict described above, and the only honest response to it is a structural one: the sprint does not sell Mediathink’s services, and the engagement ends where the doctrine says it ends. If a client wants implementation afterwards, my answer about my own firm is the same as my answer about any other — that is a separate decision, made separately, by them, and preferably with someone who was not in the room ruling on the claims.
And it means this cannot be run internally by someone reading the method, however capable they are. Not because the method is secret — it is written down, and much of it is public — but because the method depends on a position the internal operator does not occupy. Someone who is inside the company, or selling the company something else, cannot say that’s for your agency, your guru, or your astrologer and be believed. They are one of the three.