This B2B SaaS Founder Scored 61/100. Here Is What Changed in 12 Weeks.

The four issues identified were: (1) ICP scatter — the company was selling to account executives at companies ranging from 50 to 5,000 employees, and the retention data showed meaningful divergence between the two ends of that range, with the enterprise segment churning at 3x the rate of the mid-market segment; (2) a North Star metric that measured logins rather than a core workflow completion event; (3) overconfidence bias in the founder's assessment of competitive differentiation, evidenced by the gap between how the founder described the product and how churned customers described it in exit interviews; (4) an onboarding sequence built around features the product team assumed were most valuable, rather than the workflow completion data showed mattered most for retention.

The PMF score of 61 reflected that the product had clear evidence of value in a specific segment (mid-market AEs at companies with 200–1,000 employees) but was not structured to maximise or even consistently deliver that value. The risk grade of C indicated that without structural changes, the company was likely to plateau at its current revenue level rather than compound.

The diagnostic was explicit about priority. The ICP scatter and the North Star metric were the highest-priority issues because they were upstream of all other decisions — fixing them would give the team accurate data with which to address the other two issues. The onboarding and competitive positioning issues were secondary not because they were unimportant but because they couldn't be fixed correctly until the team knew exactly who they were optimising for.

First, the company narrowed the ICP to mid-market AE teams (50–200 AEs) at SaaS companies above $10M ARR, explicitly pausing outbound to enterprise accounts. This was uncomfortable. Two pipeline deals were in the enterprise range and had to be deprioritised. The founder made the decision based on the retention data, not on gut instinct.

Second, they replaced the 'logins this week' North Star with 'accounts enriched per AE per week,' which was the activity that correlated most strongly with retention in their cohort analysis. This change took 11 days of instrumentation work. The new metric showed immediately that onboarding was failing: most new AEs enriched fewer than 3 accounts in their first week, while the cohort of retained AEs who had been with the product for 90+ days enriched 12-18 per week.

Third, they rebuilt the onboarding sequence around the enrichment workflow rather than the feature tour. The previous onboarding had been a 7-step product walkthrough. The new onboarding had 3 steps focused on completing a first enrichment on a real account within 10 minutes. The 30-day activation rate increased from 34% to 61% over the 8 weeks following the change.

At week 12, the founder retook the Clarity on PMF diagnostic. The score moved from 61 to 74 — risk grade B. The ICP scatter issue was resolved. The North Star metric was now correctly instrumented and correlated with retention. The onboarding issue was substantially improved. The remaining gap was competitive differentiation: the founder's framing of the product's competitive advantage still did not match what customers cited as the primary reason they chose the product over alternatives.

The competitive positioning gap is the hardest to resolve because it requires the founder to hear something uncomfortable — that the reason customers buy the product is often not the reason the founder thinks they should buy it. In this case, customers were buying primarily because of the data quality in a specific vertical (SaaS companies with PLG motions), not because of the broader feature set the founder had emphasised. The next phase of work is repositioning around that specific advantage.

The company is now in a more useful position: it knows exactly what it is good at, for whom, and why. The 12-week plan gave the founder a structured way to act on the diagnostic rather than treating it as a report to read once and file.