Messaging That People Can Repeat Without Reading It Twice
Internal agreement and external clarity are different tests, and passing the first usually costs you the second. Write the sentence a customer will repeat, then rank everything else beneath it.
On This Page
Ask two customers what you do. If the answers do not resemble each other, the problem is not that your copy is unclear. It is that nothing in the material told them which sentence to keep.
A brand messaging framework is not a document that makes everyone in the room comfortable. It is a ranked set of claims, written so the top one survives being repeated by somebody who was half listening. Most frameworks we are handed do the opposite. They are accurate, complete, signed off, and impossible to say out loud, which is why messaging and verbal identity work usually starts by deleting rather than writing.
The failure has a signature you can spot in thirty seconds. Every stakeholder got a clause. Product got "platform". Sales got "end to end". The founder got "trusted". Legal removed the one verb carrying a promise. What comes out is defensible and inert. It reads fine in the deck and dies the moment it leaves the room, because the person repeating it has no idea which part they are allowed to drop.
So they drop all of it and improvise. That improvised sentence is your actual messaging. You simply did not write it.
The sentence a customer will actually say
The unit of messaging is not the paragraph. It is the sentence one person says to another while you are not in the room. Almost every considered purchase gets relayed at least once: to a manager who controls the budget, to a partner who has to live with the choice, to a procurement form with a small box. Your homepage does not make that trip. One sentence does, and it is rarely the one in the hero.
So write it down. Literally: the sentence you believe a customer says about you today, in their words, including the parts that are wrong. Then the sentence you want said. The distance between the two is the assignment. It is a better hour than a workshop about values, and it tends to reveal whether you have a wording problem or a positioning problem, which get treated the same and are not the same.
What makes a sentence portable
Sentences that travel share a shape.
- One claim. Two claims get repeated as one, and you do not get to choose which one survives the trip.
- A category the listener already holds. "We do X for Y" beats a coined phrase, because a coined phrase has to be explained by the person repeating it, and they will not bother.
- Words the customer used first. Borrow from sales calls and support tickets. The strategy deck is the last place to look, not the first.
- No qualifier that can be dropped without loss. If "enterprise-grade" can go, it will go, and it will take whatever credibility it was carrying with it.
- Something a competitor could not say back. A sentence three rivals would happily sign is not messaging. It is a category description with your logo on it.
That last point causes the most arguments. Teams resist it because the distinctive version is always narrower, and narrow feels like lost revenue. It is not. Narrow is the only version anyone retains. Being noticeably different in a way people can describe beats being marginally better in a way they cannot, which is the case for distinctiveness over differentiation, and it applies to sentences as much as to logos.
Why consensus copy fails outside the room
Internal agreement and external clarity are different tests, and passing the first tends to cost you the second. Consensus is reached by addition. Somebody says the sentence undersells the analytics, so a clause goes in. Somebody worries it excludes mid-market buyers, so a qualifier goes in. Nobody ever objects that a sentence has too many clauses, because nobody in that room will have to repeat it under pressure to a stranger.
The clause tax
Every clause you add lowers the odds of the sentence being retold intact, and not because people are careless. Memory keeps structure rather than text. What survives a retelling is a shape: a subject, a verb, one distinctive noun. Everything else gets rebuilt by the speaker out of their own vocabulary. That is how "we help ambitious teams accelerate their digital transformation" arrives at the other end as "they do websites, I think".
Test any candidate by deleting each clause in turn and asking whether the meaning changed. Clauses that survive deletion are decoration, and you are paying for that decoration with recall. There is no other currency in this exercise.
The room is not the market
The people approving the words hold context nobody else has. They know which product "platform" refers to, which three competitors "unlike others" is aimed at, and why the line about reliability is there at all. Strip that context away and most approved copy stops meaning anything in particular. The cheapest correction is to read the sentence to somebody who has never seen the product, then ask them what we sell and who for. If they hesitate, the sentence is not finished. A brand audit will tell you the same thing, more thoroughly and more slowly.
There is a harder version of this problem: sometimes a company cannot agree on the sentence because it cannot agree on what it sells. Our parent company has written about what gets sold when a brand stops working, and the diagnosis carries over. When a messaging exercise stalls for three weeks, it is usually reporting a real disagreement rather than a writing difficulty.
What a brand messaging framework should contain
A framework is a hierarchy, not a library. Its job is to say which claim outranks which, so that when only one sentence fits, four different people independently pick the same one. Four levels cover almost every company we work with.
| Level | What it is | Where it shows up | Test it has to pass |
|---|---|---|---|
| Core sentence | One claim, the thing you want repeated | Homepage hero, sales opener, the answer to "what do you do" | A stranger repeats it accurately an hour later |
| Proof lines | Three or four claims that make the core credible | Section headings, pitch slides, proposal summaries | Each one can be checked by somebody outside the company |
| Objection answers | The sentence for the doubt that stalls the deal | Pricing page, FAQs, the second call | It names the doubt instead of talking around it |
| Vocabulary | The words you use and the words you refuse | Everywhere, including support replies | Two writers who never met produce the same noun |
Notice what is absent: mission, values, tone adjectives, an archetype. Those can be useful internally, and none of them is messaging, because none of them is a sentence anybody says to a buyer. The same instinct produces brand guidelines nobody follows: a careful document describing a philosophy, when what a writer needed at four in the afternoon was the correct noun for the thing and permission to stop.
The vocabulary list does the most work
The least glamorous level changes output fastest. A vocabulary list has two halves: the words you use for your own things, and the words you refuse. Customer or client, pick one. Platform or service, pick one, and say what the other means if it appears. Refusals matter more than preferences, because a refusal settles an argument without a meeting, and a preference just starts one.
Keep it to a page. A vocabulary list that runs to eleven pages is a dictionary, and dictionaries get consulted by nobody working at speed. The point of this artefact is that a person writing a support reply at pace reaches for it, finds the answer in four seconds, and carries on.
The repeatability test
This takes an afternoon, costs nothing, and is the only piece of messaging research we would call non-negotiable.
- Say it once, out loud, to five people who do not work here. Friends of the team are fine. Do not send the document. Say the sentence and stop talking.
- Wait an hour. Talk about something else entirely. The delay is the test. Instant recall proves only that they have ears.
- Ask each of them to tell you what we do. Write down exactly what comes back, including the hedges and the "something about".
- Compare the five answers to each other, not to your original. Your original is not the target. Agreement between strangers is.
- Cut what everybody dropped, rewrite what nobody kept. A clause that five out of five people discarded is not being defended by anyone except you.
What you are measuring is convergence, not accuracy. Five people who get it slightly wrong in the same direction have told you something precise: the sentence has a gravity you did not intend, and the fix is one word rather than a rewrite. Five people who get it wrong in five different directions have told you the sentence has no shape at all, and no amount of polish will give it one.
Run it again after any real change to the offer. The test is cheap enough to repeat and the result decays, because the market keeps learning new words and your competitors keep teaching them.
Messaging that is still moving will move the layout with it, and hero sections get rebuilt three times for reasons that were never visual. Freeze the core sentence and the vocabulary before a designer opens a file. It is the cheapest of the decisions that should freeze before a build starts, and the one most often skipped.
Now machines repeat it as well
There is a second retelling audience and it does not attend workshops. Assistants and search features summarise a company from whatever text they can reach, and they behave a lot like a distracted human: they keep the sentence that is easiest to lift and hardest to misread. Google publishes guidance on how content appears in AI features, and the practical consequence is unglamorous. Be consistent, be specific, and put the plain description somewhere it can actually be read.
Consistency matters more here than cleverness. A company described one way on the homepage, a second way on the about page and a third way in the footer gives a summariser nothing stable to hold, so it produces an average of the three, and averages are never distinctive. Publishing the same short description in your organisation markup, your about page and your profile listings is dull work with a compounding return.
Where a person or a very small team is the brand, the same rule applies to whatever profile you point people at. A product like Nichevio assembles a mobile-first profile site out of structured widgets, and the structure is the useful part: a field for the one-line description forces the sentence to exist rather than letting it hide inside a paragraph nobody finishes.
Where the hierarchy breaks in practice
Frameworks rarely fail on the day they are written. They fail six months later, quietly, in about five recognisable ways.
- A feature ships and somebody edits the core sentence to mention it. The sentence gains a noun and loses its shape. That edit is usually a symptom of the roadmap and the position drifting apart, which is the separate question of when a feature request is really a positioning problem.
- A second audience appears and the sentence is widened to cover both. Two audiences need two sentences and one hierarchy. They never need one sentence joined by "and".
- A competitor launches something loud and the sentence quietly becomes a rebuttal. Rebuttals do not travel. Nobody relays your argument on your behalf.
- A campaign line outranks the core sentence in reach while saying something slightly different. Within a quarter the market believes the campaign, and the framework is a historical document.
- Somebody in sales writes a better sentence than yours and nobody writes it down. This is the good failure, and it is the one worth going to look for on purpose.
If nobody can say the sentence without the document open, the document is the deliverable and the message does not exist.
What this does not fix
Repeatability is a constraint, not a strategy. A memorable sentence about a position nobody wants will travel efficiently to the wrong conclusion, and the efficiency is no consolation. If the underlying offer is undifferentiated, sentence craft will feel like it is failing when it is actually reporting. Listen to the report rather than commissioning a second round of copy.
Some categories genuinely need two sentences. Regulated products with a compliance buyer and a user buyer. Technical infrastructure where the person who signs and the person who suffers are different people. In those cases write both, rank them, and be explicit about which room each belongs in. What you must not do is average them into one sentence that satisfies neither buyer and confuses both.
Messaging problems also get misdiagnosed as identity problems, constantly. A new logo will not make a sentence repeatable, and a rebrand undertaken to fix wording is an expensive way to buy a workshop. The question of how you know it is time to rebrand has real answers, and "our messaging is muddled" is not among them.
Where we would start on Monday
Pull twenty sentences you did not write. Sales call notes, support tickets, review text, the line a partner uses when introducing you, the description a customer put in their own material. Sort them by how often the same idea repeats. You are looking for the claim that already travels, because promoting an existing claim is far cheaper than installing a new one, and there is nearly always one available.
Then write three candidate core sentences and run the repeatability test on all three in the same afternoon, using different people for each. Do not tell the testers which one you wrote last or which one you prefer. Pick the sentence that comes back most intact, not the sentence you most enjoy reading. Those are different sentences more often than anyone expects.
Last, write the vocabulary list, both halves, and put it where writing actually happens rather than in a slide deck nobody opens. If the framework survives its first month of real pressure, it will be because somebody could reach for it mid-email. If you want a second pair of eyes on a sentence that is not landing, tell us what your customers say back to you and we will start there rather than with a blank page.
Common questions.
What is a brand messaging framework?
A brand messaging framework is a ranked set of claims that tells everyone in a company which sentence outranks which. At minimum it holds one core sentence, three or four proof lines, prepared answers to the objections that stall deals, and a vocabulary list covering the words you use and the words you refuse. Its purpose is consistency under pressure, not completeness.
How long should a positioning statement be?
One sentence with a single claim in it. Length matters less than clause count, because each additional clause reduces the chance of the sentence being repeated accurately by somebody who heard it once. If you cannot say it in a breath, or if a listener can drop half of it without changing the meaning, it is carrying decoration rather than argument.
How do I test whether messaging is memorable?
Say it once to five people outside the company, wait an hour, then ask each of them to describe what you do. Record their exact words. Compare the five versions to each other rather than to your original, because agreement between strangers is the real signal. Convergent errors point to a single word that needs changing; scattered answers mean the sentence has no shape.
Why does messaging that everyone approved still confuse customers?
Because the approvers hold context the customer does not. People in the room know which product a generic noun refers to and which competitor a comparison targets, so the sentence reads as specific to them and as noise to everybody else. Approval by addition makes it worse, since every stakeholder adds a clause and nobody is responsible for removing one.
Should messaging be finalised before a website redesign?
Yes, at least the core sentence and the vocabulary list. Layout decisions follow from what the words are trying to do, so unresolved messaging turns into repeated design work on hero sections, navigation labels and page order. Freezing the top of the hierarchy before design starts is cheap; discovering halfway through a build that the claim has changed is not.
How often should a messaging framework be revisited?
Review it whenever the offer changes materially, a new buyer type appears, or a campaign line starts outranking the core sentence in reach. Otherwise an annual check is enough. Frameworks decay because the market learns new vocabulary and competitors keep teaching it, so the repeatability test is worth rerunning rather than assuming last year result still holds.
Facing this in your
own business?
Tell us where you’re headed — we’ll map the shortest honest route.