B2B positioning: turn customer evidence into a clear message
Positioning decides who you are for, against which alternative, why you are better and what proves it. Messaging comes after. A worked example from generic headline to defensible claim, and where AI helps.
The short version
- Positioning is a set of decisions; messaging expresses them. Fix the decisions before arguing about words.
- Start from what customers would do without you, including spreadsheets, hiring and doing nothing.
- Link each difference to a consequence, a best-fit buyer and proof you can show today.
- Use AI to draft and stress-test wording inside those constraints, never to decide positioning or stand in for buyer research.
On this page
Most messaging meetings are arguments about words. Should the headline say “platform” or “tool”? Is “effortless” too much? The disagreement usually is not about words at all. It is about decisions nobody has made yet: who the product is for, what it is being compared with, and why a buyer should believe it is better.
Positioning is that set of decisions. It defines which customers you serve best, compared with which alternative, because of which meaningful difference, backed by which proof. Messaging is how you express those decisions to a particular audience. Make the decisions first, from customer evidence. Use AI afterwards, to test and draft wording inside the constraints you have set.
Positioning, value proposition, messaging and tagline
These four terms describe different layers. Keeping them apart stops a team from trying to fix a strategy problem with a better adjective.
| Layer | What it is | Who uses it | Example (fictional reporting tool) |
|---|---|---|---|
| Positioning | The strategic choice: for whom, against what, why better, with what proof | Internal: leadership, product, marketing, sales | For operations teams reporting from several systems without a data engineer; instead of rebuilding spreadsheets each week |
| Value proposition | The specific value a type of customer gets | Sales conversations, key pages | Weekly reports assembled from existing systems, so the team reviews numbers instead of compiling them |
| Messaging | How the value is expressed for each audience and stage | Website, ads, emails, decks | For the operations director: "Your Monday report, ready before your Monday meeting" |
| Tagline | The shortest memorable expression | Brand surfaces | Optional; often unnecessary for a B2B product |
You can have clear positioning and no tagline. You cannot have useful messaging without positioning, because every message quietly assumes a buyer and an alternative.
Start with what the customer would do without you
The most useful positioning question I know comes from April Dunford’s book Obviously Awesome (2019): what would your customers do if your product did not exist? Her approach builds positioning from these competitive alternatives, the attributes that make you different from them, the value those attributes create, and the customers who care most about that value.
The answer is rarely just a list of competitors. For a B2B product it usually includes:
- Doing it manually: spreadsheets, copy and paste, a weekly ritual someone resents.
- A broader platform: something the company already owns that could do the job, badly or with effort.
- Hiring: an analyst, an agency or a contractor.
- Doing nothing: living with the problem because changing feels riskier than the pain.
“Doing nothing” is the alternative teams forget, yet a deal lost to “no decision” is still a lost deal. It changes the message. If the main rival is a spreadsheet, claiming you have “more features than Competitor X” misses the point. The buyer needs a reason to change at all.
Where does the evidence come from? From the same records you used for your ideal customer profile: lost deals, sales call notes, and customers’ own descriptions of what they used before.
Connect a difference to value and to the right buyer
A difference only matters if it creates value for someone. Work through a short chain for each candidate difference:
- Capability: what can you do that the alternative cannot, or cannot do easily?
- Consequence: what does that allow the customer to do or avoid?
- Best-fit buyer: who cares most about that consequence?
- Evidence: what would convince a sceptical buyer that the claim is true?
A capability without a consequence is a feature list. A consequence that every alternative also delivers is not a difference. A difference that no buyer cares about much is trivia. The positioning lives where all four links hold.
Build a claim and proof table
Write down every claim your current messaging makes, then add the proof you could show a buyer today. This step is uncomfortable and useful.
Two tests help:
- The swap test. Could a competitor put this exact sentence on their homepage? If yes, it is not positioning. “Powerful, intuitive and built for growing teams” passes for almost any software company.
- The comparison test. Would this claim survive a side-by-side comparison with the buyer’s actual alternative? “Faster” than what, measured how?
Claims that fail both tests are either removed or rewritten into something specific enough to prove.
A worked example: from generic to defensible
This example is illustrative. The product, alternatives and claims are invented, and none of the claims below should be published without the evidence listed beside them.
A fictional company sells a reporting tool to operations teams. Its buyers currently use one of two alternatives: spreadsheets rebuilt every week, or a broad analytics platform the company already licenses but that needs a specialist to configure.
The current homepage headline reads:
The powerful, intuitive reporting platform that helps teams make better decisions faster.
Every word could appear on any competitor’s site. “Powerful” compared with what? “Better decisions” is a hope, not a claim. The sentence does not say who it is for or what it replaces.
Working through the chain changes the claims:
| Alternative | Meaningful difference | Buyer value | Best-fit situation | Evidence needed before publishing | Message |
|---|---|---|---|---|---|
| Weekly spreadsheets | Pulls from existing systems on a schedule | Stops the weekly rebuild and copy-paste errors | Teams reporting from several systems | Live demo on a buyer's own sources; list of supported systems | Your weekly operations report, assembled from the systems you already use |
| Broad analytics platform | Usable without a data engineer | No waiting for specialist time to change a report | Teams without dedicated data staff | Setup walkthrough; customer quote used with permission | Built for operations teams that do not have a data engineer |
| Doing nothing | Low-risk start with existing data | Change without a migration project | Teams wary of a large implementation | Documented onboarding steps and typical scope | Start with one report and the data you already have |
The revised headline might read:
Your weekly operations report, assembled from the systems you already use. Built for teams without a data engineer.
It is less impressive and far more useful. It names the job, the alternative it replaces and the buyer it suits. A competitor aimed at data teams could not copy it without contradicting its own positioning.
Use AI for wording, inside the constraints
Once the claim and proof table exists, AI becomes genuinely helpful. Give it the table and ask for wording options, with clear limits:
Two other uses work well. Ask the model to act as a sceptical buyer and list the objections each claim invites. And ask it to rewrite a message for a different member of the buying committee, such as finance instead of operations, without changing the underlying claims.
What AI cannot do is decide your positioning or tell you whether buyers prefer one message. A model that “prefers” option B is reporting a pattern in text, not the reaction of your market. Competitive facts it states from memory may be out of date or wrong; gather them properly, as described in how to do competitor analysis with AI.
Test the message with real buyers
The cheapest useful test is comprehension. Show the message to a few people who match your ideal customer profile and ask them to explain, in their own words, what the product does, who it is for and what they would use instead. If their answer does not match your positioning, the message has not worked, however good it sounds internally.
Relevance is the next test: does the buyer recognise the problem as theirs, and would it change what they do next? Sales calls are a natural place to try variations. When you have enough traffic, a structured test on a landing page can compare messages; designing a marketing experiment covers how to set one up so the result means something.
What if competitors make the same claims?
Then those claims are table stakes, not positioning. Keep them where buyers expect them, and move your emphasis to the difference only you can prove. If you cannot find one, that is a real finding: the answer may be to narrow the target market until a difference matters, or to change the product. No wording fixes a missing difference.
Blank positioning worksheet
Copy this into a document. Fill one row per competitive alternative, and do not fill the message column until the evidence column is honest.
| Alternative | Meaningful difference | Buyer value | Best-fit situation | Evidence (available or needed) | Message |
|---|---|---|---|---|---|
| [What the customer would do instead] | [What you do that it does not] | [What that allows the buyer to do] | [Who cares most] | [What you can show today] | [How you say it] |
Once the positioning is settled, it shapes what you publish. The next step is to map content to the questions buyers ask on the way to a decision: B2B content strategy: map content to buying decisions.
Questions worth asking
How is positioning different from messaging?
Positioning is the strategic choice of target customer, competitive alternative, meaningful difference and proof. Messaging is how you express that choice to a particular audience at a particular stage. One positioning usually produces several messages, for different roles in the buying committee.
Can AI decide our positioning?
No. It has no access to your deals, customers or the reasons buyers chose you, and its preference between two messages is not a market response. It is useful after the decisions exist: drafting options within your claim and proof table, listing likely objections and adapting a message for a different role.
What if competitors make the same claims?
Then those claims are table stakes. Keep them where buyers expect them and move your emphasis to a difference only you can prove. If there is none, consider narrowing the target market until a difference matters, or changing the product. Wording cannot create a difference that does not exist.
