The situation
A B2B SaaS company with real traction. Acquisition is doing its job. Traffic is up year on year, the cost of it is stable, and the top of the funnel looks like the top of a healthy funnel. Conversion from visit to trial is poor, and has been poor for as long as anybody has been measuring it.
The homepage is accurate. It explains what the software does, lists the capabilities, and could be pasted onto a competitor domain with two words changed. It does not say who the software is for, or what is different for that person once they have bought it.
What the company usually believes
That it is a traffic quality problem, or a conversion rate problem, or both. By the time we see it the team has usually run a redesign, tested headlines, moved the call to action above the fold and added social proof. Each change moves the number by less than the noise in the measurement.
Which of the three it is
A positioning problem.
How the other two get ruled out
Not a downstream leak. The accounts that do convert stay at a normal rate and the retention curve flattens where a healthy one flattens. The bucket holds. Whatever is wrong sits upstream of it.
Not a lack of urgency. Deals that reach a conversation close at a reasonable rate and on a reasonable clock. Buyers who understand the offer act on it. Very few buyers understand the offer.
The evidence
Three things, none of which needed new instrumentation.
Ask five people inside the company who the product is for. Five answers, all reasonable, none the same. A company cannot say to the market what it has not said to itself.
Read the homepage as somebody who has the problem but does not know the category exists. Nothing on the page tells that person the page is about them.
Sort the existing customers by whether they renew without being chased. The ones that do have something in common, and it is usually not the segment the marketing addresses.
What we do
We find the position in the customer base rather than inventing one. The accounts that stay are the argument. We name that buyer specifically enough that the definition excludes people, write the claim in terms of what changes for them rather than what the product does, and build the message architecture that carries the same claim through the site, the onboarding, the sales conversation and the paid copy.
What changes
The claim made at the top of the funnel and the promise the product keeps at the bottom of it become the same claim. A visitor who fits can tell within a few seconds that the page is about them. A visitor who does not can tell that too, which is the more valuable half.
What the company stops doing
Buying traffic against terms that describe the category rather than the buyer. Testing headline variants on a claim that is the wrong claim in every variant.
The lesson
More traffic does not fix unclear positioning. Every extra visitor is money spent sending a stranger toward a claim they cannot place themselves inside, so the cost of the confusion scales with the spend.