Your trial is not broken: you are bringing in the wrong users

When a trial converts poorly, the common reaction is to inspect onboarding. Teams change screens, tooltips, emails, in-app messages and reminder sequences. The work feels sensible because it sits close to the visible problem: people enter, they do not pay, so something in the trial must be broken.

On July 7, 2026, Ben Williams published a PLG.news piece titled Stop fixing your trial. Fix who’s in it. The thesis is simple and uncomfortable: often the problem is not the trial, but who enters it.

For a product team, that distinction changes everything. If signups are not qualified, you can improve onboarding for months and still work on the wrong constraint. The funnel shows a drop-off, but it does not automatically tell you where the problem starts.

Signups are too convenient as a metric

Signups are easy to measure and easy to move. A sharper campaign, a clearer landing page, an SEO article that catches more traffic, a post that performs well. The number goes up and it feels like progress.

But not all signups mean the same thing. Some come from people with the right problem, budget, context and urgency. Others are curious visitors, students, competitors, people outside the target, users who misunderstood the promise or companies too small to buy.

If you put them all in the same chart, the average becomes a polite lie. It tells you conversion is low, but not whether you are talking to potential customers or to people who were never going to buy anything.

This is the same diagnostic discipline behind discovery and product prioritization. Before optimizing the product surface, the team has to understand which problem, segment and evidence actually deserve the next cycle.

The question is not only activation

Many PLG playbooks assume that the user entering the trial is already more or less qualified. From there, it makes sense to work on activation, time to value, lifecycle, experiments and nudges. But if that assumption is false, the playbook answers a different question.

Better onboarding can help the right user understand value sooner. It cannot turn the wrong user into a customer. It can make the path smoother, but it cannot change the fact that the person has no real problem, no budget or no authority.

That is why the first discovery should not look only inside the product. It should go upstream: where signups come from, what people expected, which promise they read, which job they thought they were solving, why they decided to try.

Manual work becomes useful again

When the numbers are still manageable, the fastest way to understand what is happening is not always a more refined query. Often it is taking two weeks of signups and looking at them one by one.

Who are they? Real company or personal account? Role consistent with the ICP? Plausible problem? Company domain? Signs of urgency? Did they use a feature that shows real intent, or did they click through two screens and leave?

This work does not scale well, and that is exactly the point. If you have hundreds of signups, you can afford a manual review. You cannot afford months of optimization on a funnel that may be measuring curiosity rather than demand.

Where a Fractional PM fits

This is where the role of a Fractional PM is not “fix the roadmap”. It is helping the team decide which problem it is actually solving.

If a founder sees growing signups and low conversion, asking for more activation experiments is natural. An external PM needs to be able to ask the less comfortable question: are we sure we are bringing in the right people?

From there, priorities change. Before rebuilding onboarding, maybe positioning needs to be clearer. Before adding emails, maybe acquisition needs to change. Before building a feature to improve activation, maybe the team needs to separate segments and understand which users deserve product, sales or content.

Product culture protects the diagnosis

A good product culture is not visible only in the number of discovery calls a team runs. It shows up in how strongly the team protects the diagnosis before jumping to the solution.

Optimizing a trial is legitimate work. The problem starts when it becomes the automatic answer to any low metric. At that point, the team is using the product to fix a problem that may belong to market, positioning, channel or pricing.

The roadmap question is not “how do we increase trial conversion?”. It is more precise: which users do we actually want in the trial, which signals tell us they are qualified, and which part of the funnel is bringing in people who should not be there?

If that answer is missing, onboarding may still improve. But the team will keep polishing a door that opens onto the wrong people.