A value proposition rests on three components: a named segment, a problem dated by a trigger, and a measurable result. It is not validated internally but across five interviews, with a single test: can the customer restate it in their own words? Four phrasings disqualify it outright, and the first is the one that describes the product. Budget two weeks, only one of them writing.
Most value propositions we inherit were written in a single morning, internally, around a whiteboard. They are elegant, everyone agreed, and they serve no purpose: no salesperson uses them, no prospect remembers them.
This is not a writing talent problem. It is a method problem: a value proposition is not written, it is discovered and then formulated. Here is how we proceed, and the single test that separates the ones that hold from the ones that decorate.
What it is, and what it is not
A value proposition is not a slogan, a product description, or a sales argument. It is a sentence stating, from the customer’s point of view, what changes for them.
It differs from positioning, which is the upstream strategic decision: which place you occupy for a segment, relative to the alternatives available to them. The value proposition is the testable formulation of that decision. So it cannot be written until positioning is settled, and that is the first cause of failure: teams hunting for a sentence when what they lack is a choice. We cover that decision in positioning: say who you serve.
The three components
A named segment. Not “B2B companies”, but a set recognisable from the outside: size, sector, model, maturity. The test is simple: someone in the segment should be able to think “that is me” in a second, and someone outside it should think “that is not me” just as fast. The second half matters as much as the first.
A problem dated by a trigger. A permanent problem does not make anyone buy. What makes people buy is a problem that becomes urgent at an identifiable moment: a funding round, a hire, a regulatory obligation, a growth threshold crossed. Naming that moment turns a general promise into an actionable one, and it is what later makes prospecting on observable signals possible, as we describe in signal-based selling.
A measurable result. Not a capability, a result. “A platform that centralises your data” is a capability. “Your reps stop losing two hours a day hunting for information” is a result. The difference is not stylistic: the first forces the prospect to do the arithmetic of what they gain, and most will not.
The test protocol
This is the part almost nobody runs, and it is the part that decides.
Five interviews, not a survey. Take five people from the target segment, at least two of whom are not customers. Read them the proposition once, then ask them to restate it in their own words.
The only criterion that matters: the restatement. If they manage it and the meaning stays faithful, it holds. If they find it “interesting”, it says nothing: that is the polite answer of someone who retained nothing. If they ask you to clarify, a component is missing, and their question usually tells you which one.
Never test internally. Your team will always understand the proposition, because they already know the product and the context. An internal test only measures consistency with what everyone already knows, which is precisely the blind spot.
Write down the words they use. The final version of your proposition will come out of those five interviews, not out of your workshop. It is the same principle as in diagnosis: the customer’s words beat yours, and they convert better.
The four phrasings that disqualify
The one that describes the product. “A collaborative management SaaS solution” could name a thousand companies. If your sentence survives having your name swapped for a competitor’s, it says nothing about you.
The one promising a universal benefit. Saving time, cutting costs, improving collaboration. Those promises are true, shared by your whole market, and therefore inert. They reassure internally because nobody can contest them.
The one stacking two segments. “For SMBs and large accounts” means the choice was not made. The sentence widens until it touches nobody, and that is the symptom of an unsettled ICP, the subject of why an ICP that is too broad kills growth.
The one naming the technology. Putting artificial intelligence or a particular architecture at the centre of the promise moves the conversation to the how. The customer does not buy a mechanism, they buy a result, and technology becomes proof, not the way in.
What varies, what stays
As soon as you serve two segments whose main problem differs, one proposition stops being enough. Trying to write one broad enough to cover both mechanically produces an empty sentence.
What stays common is the company narrative: what you believe, why you exist, what you refuse to do. What varies is the promise made to each segment, and then the message addressed to each role inside that segment. Those are three distinct layers, and confusing them produces either corporate language nobody remembers, or a mosaic of promises with no spine.
When to rewrite it
Three signals, one of which nobody watches.
The priority segment changes: the proposition follows, mechanically. The market’s alternatives change: what used to differentiate you becomes a standard, and your promise loses its edge without anything moving on your side.
The third is the most reliable and the least heard: your salespeople stop using it of their own accord. They fall back on their own phrasing, the one that works in meetings. That is not indiscipline, it is information. Listen to ten recorded calls and compare what actually gets said with what is written on the site. The gap contains the next version, and it is also the raw material of a usable playbook, as we explain in a sales playbook that survives your best closers.
Where to start
Do not open a blank document. Take your last five customers who get a measurable result and call them with a single question: “what problem did you have, the day you contacted us?”
Their answers contain the three components. Your job is no longer to invent a sentence, it is to pick the one that recurs, then go and test it on five people who do not know you.
Frequently asked questions
- What is the difference between positioning and a value proposition?
- Positioning is the place you occupy in a segment's mind, relative to the alternatives. The value proposition is the sentence expressing that place from the customer's point of view. The first is a strategic decision, the second its testable formulation. You cannot write the second until the first is settled.
- How do you test a value proposition?
- Read it to five customers in the target segment and ask them to restate it in their own words. If they can and the meaning stays faithful, it holds. If they find it 'interesting' or ask you to clarify, it says nothing. The test is never run internally: your team will always understand it, because they already know the product.
- Do you need one value proposition per segment?
- Yes, as soon as you address two segments whose main problem differs. A single proposition broad enough to cover both will speak to neither. What stays common is the company narrative; what varies is the promise made to each segment.
- How often should it be rewritten?
- When the priority segment changes, when the market's alternatives change, or when your salespeople stop using it of their own accord. That last signal is the most reliable: a proposition the field has abandoned is a proposition that no longer works, whatever the committee that approved it says.