the-leak-finder

Installation
SKILL.md

Map both funnels before optimising either. Read The Product-Qualified Path, and Why MQL Alone Is the Wrong Model in references/funnel-benchmarks.md.

  • Ask whether any self-serve path exists before assuming a single sales-led funnel. Most companies with a signup form are running two funnels and measuring one.
  • Sales-led qualifies on MQL, product-led on PQL — a PQL has used the product and shown buying behaviour, an MQL downloaded something. Bare logins never qualify: a habitual logger with no expansion behaviour is a habitual user, and those are disproportionately the accounts quietly evaluating alternatives.
  • Where both paths run, compare them. A measured case showed MQL→SQL of 10.9% against PQL→SQL of 57.9% at the same company — a 5.3x gap at the qualifying step, where the leverage is routing traffic into the product path rather than repairing the MQL path. That conclusion is invisible if only one funnel is mapped.
  • MQL→SQL is a distribution, not a floor: 13% cross-industry median, 18-22% B2B SaaS, 35-40% top quartile, and 39-40% with behavioural scoring — roughly triple the median, which is the same idea as a PQL applied to the sales-led path. That is usually the recommendation, not more nurture.
  • A blended qualifying rate cannot be acted on. SEO converts to SQL at ~51%, PPC ~26%, webinar ~17.8%. Splitting by channel is the first deliverable, not a refinement.
Installs
5
GitHub Stars
9
First Seen
Jul 26, 2026