In short
- The four versions are not carelessness. Each team is solving a different problem with the same raw material.
- Nothing looks broken while the founder is in every conversation. The drift surfaces the quarter you stop being in them.
- A document does not hold four teams together, because a document written for one audience has to be rewritten for the next three, and the rewrite is where drift enters.
- The diagnostic costs nothing: ask four people from four teams the same question separately, and compare what comes back.
- This is a different problem from a pitch that does not land. Fixing the wrong one of the two is the expensive mistake.
Why does every team end up explaining the product differently?
Because each one is answering a different question, and nobody wrote down the answer they should all be starting from.
Sales is trying to win a specific meeting with a specific sceptic, so the explanation bends toward whatever that person pushed back on last time. Marketing is trying to be found by strangers. The same thing gets flattened into something that reads well cold. Delivery is trying to stop a new customer expecting the wrong outcome, so it leads with the caveats. Product is defending a roadmap, so it leads with what is coming.
So none of that is sloppiness. Each version is locally correct, built by someone doing their actual job well. But four locally correct versions add up to a company that appears to sell four different things, and any buyer who speaks to more than one of your people notices.
The sales and marketing half of this has been documented for twenty years, and it is usually framed as a turf problem between two functions. But in a complex B2B company it is rarely turf. It is that both functions are reconstructing an argument from memory, because the argument was never written down in a form either of them could adapt.
What does it actually cost when four teams disagree?
Three things, and only the first one is obvious.
The obvious cost is buyer confusion. A prospect reads the website, takes a sales call, then talks to someone in delivery during a reference check, and assembles three partial pictures into a worse one. That is the cost people notice, because it shows up as a stalled deal.
The second cost is that nobody can tell which version is working. If four teams are running four arguments, a won deal does not tell you what won it, and a lost one does not tell you what failed. You lose the ability to learn from your own pipeline, which is expensive in a way that never appears on a dashboard.
But the third is the one that compounds. Every new hire learns whichever version their own team happens to be running, so the four versions harden into four local traditions. A few years in, changing the explanation means changing four things that no longer agree on what they are a version of.
A diagnostic that costs nothing.Ask four people, one from each team, separately and in writing, why a customer should choose you over the obvious alternative. Do not prompt them and do not let them confer. Then read the four answers next to each other. Most founders find the result more uncomfortable than they expected, which is the point of running it.
What does a shared explanation have to do to survive four teams?
It has to be adaptable without being rewritten. That single requirement rules out most of what companies actually produce.
A message document fails it. Written for one audience, it has to be rewritten for the next three, and every rewrite is a chance to drop the hard part. And a slide deck fails it for the same reason, with the added problem that decks get forked and the forks never come back. A one-line positioning statement passes the adaptability test but carries too little to be useful: four teams can all agree on it and still explain the product four ways underneath.
Scroll the table sideways to see every column.
| Team | What they need it for | How their version drifts | What tells you it drifted |
|---|---|---|---|
| Sales | Winning a specific meeting with a specific sceptic | Bends toward the last objection that cost them a deal | Two reps describe the same product to two buyers differently |
| Marketing | Being found and understood by people who never speak to you | Flattens the reasoning into something that reads well cold | The site and the sales call feel like different companies |
| Delivery | Stopping a new customer from expecting the wrong outcome | Leads with the caveats and undersells the case | Handover calls open by walking back what was sold |
| Product | Defending a roadmap and choosing what to build next | Leads with what is coming rather than what exists | Roadmap items get sold before they are built |
So what passes the test is a model: the reasoning drawn as a structure rather than written as prose. A tree, a matrix, a sequence. The structure holds what has to stay constant, which is which input leads to which outcome, and leaves each team free to change the emphasis, the vocabulary and the proof they reach for. Marketing can lead with the outcome and sales can lead with the objection, and both are still visibly working from the same thing.
How do you build one that four teams will actually use?
Start from won deals rather than from a workshop, and involve all four teams in the reading rather than the writing.
- Run the four-answer diagnostic first, and keep the four answers. They tell you where the divergence actually is, which is almost never where you assumed.
- Take three recent wins and reconstruct what convinced the buyer: the problem you led with, the proof that landed, the objection that turned it, what they had to believe before they moved.
- Draw that reasoning as a structure, not a paragraph. If it cannot be drawn, it is not yet clear enough to hand to four teams.
- Give the draft to one person from each team and ask them to use it in their own next real conversation, unaided. Watch where each of them hesitates or adds something back.
- Fix what all four hesitated over. Where only one team needed something extra, that belongs in their layer, not in the shared model.
But the last step is the one that gets skipped, and skipping it is how you end up with a model that only the team who commissioned it uses.
At Strategem this is the alignment half of the work. It sits underneath a product that takes too long to explain, and we do it for a champion who has to explain it internally, for an expert whose exceptions all come back to them, and for a founder the team keeps pulling into the same sales moment.
What goes wrong when teams try to align on their own?
Usually one of three things, and each one produces a document that everybody nods at and nobody uses.
- Alignment by meeting. Four teams negotiate wording until nothing is left that any of them objects to, which is also nothing that would persuade a buyer.
- Alignment by decree. One team writes it and the others are told to adopt it. The other three keep their own version for real conversations and use the official one in reporting.
- Alignment by tooling. The model goes into a wiki or an enablement platform and the work is declared done, without anybody having tested whether a person can hold it up in a live meeting.
But all three share a root cause. The teams were asked to agree on words, when the thing they needed to agree on was the reasoning underneath the words. And agreeing on reasoning is harder to convene and much harder to fake, which is exactly why it holds afterwards.
Is this the same problem as a pitch that does not land?
No, and telling them apart is the whole reason to run the diagnostic before you spend anything.
If your four answers come back saying broadly the same thing, and buyers still do not follow you, the explanation itself is the problem. It is coherent inside the building and it is not landing outside it. That is a messaging problem, and it is the one the product messaging work exists for.
But if the four answers come back different, you cannot yet know whether your explanation lands, because you do not have one explanation to test. Fix the alignment first and the messaging question becomes answerable. Do it the other way round and you will improve one of the four versions while the other three carry on.
There is a related piece here on why buyers cannot repeat your value proposition, which is the same failure one step further out, once the drift has left the building.