Find Your Biggest Drop-Off.

A funnel analyzer turns a sequence of stage-to-stage user counts into step-by-step conversion rates plus an end-to-end rate. Then it surfaces the biggest single drop-off: the step where you're leaking the most users. That step is usually where the highest-leverage growth experiment lives. This tool works for any funnel: marketing site to signup, signup to activation, free trial to paid, demo to closed-won, anything.

The most-cited frame is Dave McClure's AARRR (Pirate Metrics): Acquisition → Activation → Retention → Revenue → Referral. But Brian Balfour argues funnels are the wrong shape entirely (growth loops are the new funnels) because they treat acquisition as one-shot rather than continuously fed by the system's own outputs. Both lenses are useful: this analyzer maps the linear stage transitions; the growth engine framework reasons about the loops connecting them.

A 50% drop is healthy at the top of the funnel and catastrophic at the bottom. A 95% drop at the awareness-to-visit step is fine; the same 95% drop at trial-to-paid means your positioning is wrong, your activation is broken, or your pricing is mismatched. Look for the step where the rate is worse than the median for your stage AND where the absolute user loss is large enough to move the topline if you fix it.

For B2B SaaS, rough public benchmarks: site-to-signup 2-5%, signup-to-activation 30-50%, activation-to-paid 5-15%, paid-to-retained-12mo 70-90%. For developer-led PLG, the signup-to-activation rate is the leverage point; for sales-led, it's opp-to-closed-won.

Pick the step with the worst leverage-adjusted leakage and run three to five experiments on it before touching anything else. Per Sean Ellis's ICE framework, prioritize experiments by Impact × Confidence × Ease, and only move on once you've compounded a 20-50% lift in that stage. Spreading experiments across all stages dilutes the signal and the lift.

Funnel analysis breaks down when your user journey is non- linear (e.g. multi-product platforms, self-serve+sales-led hybrids, communities where the "funnel" loops back). In those cases use cohort retention and loop analysis instead. The cohort analysis tool is the right next step if funnels are too thin a model for your business.