The free trial that was selecting for the wrong user
Trial signups were up forty percent. Conversion was down sixty. The funnel was working — for the wrong people.
A SaaS company I worked with had a trial signup growth of forty percent quarter-over-quarter. The marketing team was celebrating. The conversion rate had quietly dropped by sixty percent over the same period. Nobody was celebrating the second number because nobody had been asked to.
The trial was working. The trial was selecting for the wrong user. The two were not in conflict — they were the same fact stated twice.
A free trial is not a top-of-funnel widget. It is a selector. Every choice in its design — the length, the gating, the credit-card requirement, the onboarding sequence, the welcome email tone — is a filter that decides which kind of user makes it through to paid. Most teams design the trial to maximize signups, which is the easiest metric to celebrate and the most misleading to optimize. Tuning for signups inevitably means lowering the activation bar, which means filtering in more users who will never convert, which means the paid conversion rate drops on a delay long enough that the marketing team has already shipped six more campaigns by the time the drop is visible.
The diagnostic is straightforward and almost no one runs it. Take the trial signups from twelve months ago. Split them into the cohort that converted to paid and the cohort that didn't. Look at every signal you have on day one — referral source, company size, role, time-on-site before signup, first action in the product. The pattern in the converted cohort is your actual ICP. The pattern in the unconverted cohort is what the marketing team has been spending money to attract. The two patterns are usually different in ways nobody on the team has admitted out loud.
The fix is to redesign the trial against the converted-cohort pattern, not the signup-volume goal. This almost always involves adding friction — a longer onboarding question, a required integration, a credit card up front for some segment, a sales-assisted path for buyers above a threshold. Volume drops in the short term. Conversion rate climbs. Revenue per signup, the metric that actually matters, recovers within two quarters.
The harder version of this is positioning. A trial selecting for the wrong user is often the symptom of a marketing site positioned for the wrong user, ranking for the wrong keywords, running ads against the wrong jobs-to-be-done. The trial is the last gate, and gates downstream of misaligned upstream traffic cannot fix the alignment. Fix the upstream message, then redesign the trial.
The trial is not a marketing surface. It is a hiring decision the company is making about its customer base. Every signup is a candidate. The trial is the interview. Most companies run the interview to maximize candidate flow rather than candidate quality, and then wonder why the resulting cohort is so hard to retain. The cohort is hard to retain because you hired them for the wrong job.