- What's the formula for funnel conversion?
- Overall conversion = users at the last stage ÷ users at the first stage. Per-step conversion = users at stage N ÷ users at stage N-1. Drop-off between stages = users at stage N-1 minus users at stage N. Overall conversion also equals the product of every per-step rate: if your three steps are 12%, 50%, and 30%, overall = 0.12 × 0.50 × 0.30 = 1.8%. That multiplicative structure is why even small improvements at one step compound through the whole funnel.
- How many stages should my funnel have?
- As few as you can while still seeing the actions that actually predict downstream behavior. Most product funnels work well at 3–5 stages: visit, key activation event, primary value moment, paid conversion. Too few stages and you can't see where users leak; too many and noise drowns the signal. A good rule: if two stages always convert at >95% to each other, collapse them. If a single stage has a >50% drop, split it into the sub-actions and look closer.
- What's a 'good' funnel conversion rate?
- Wildly dependent on your funnel and category. For reference (Baymard, FirstPageSage, and industry surveys): e-commerce site visits to purchase typically run 1.5–3%; B2B SaaS visit-to-trial 2–5%, trial-to-paid 15–25%; mobile app install-to-day-7-retention 25–40%; email signup-to-newsletter-open 25–40%. The right comparison is not 'industry average' but your funnel last month — moving 1.8% to 2.2% is a 22% lift in business, which is huge.
- Should I focus on the worst-rate step or the highest drop-off-count step?
- Drop-off count first, rate second. A 90% drop on a stage with 50 users loses you 45 users; a 10% drop on a stage with 10,000 users loses you 1,000. The 10% step has a far bigger absolute business impact even though its rate looks healthy. That said, very-low-rate steps often signal a broken funnel design (a confusing form, a price reveal, a permissions request) and fixing them sometimes returns 2–5x gains, not 10% gains. Look at both numbers — the calculator shows both.
- Why are my counts going up between stages?
- They shouldn't be. A funnel is monotonically non-increasing — once a user drops, they're gone from this funnel. If your analytics shows a downstream stage with more users than an upstream one, something is wrong: you're counting different cohorts at different stages (window length differs), you're double-counting (a user can enter the lower stage from multiple entry points), or your tracking events fire on different conditions. Fix the data before trusting the funnel. The calculator refuses to compute when a stage exceeds the one before it.
- Can I use this for a sales funnel (leads → MQL → SQL → closed-won)?
- Yes — sales funnels are exactly the use case. Common stages: Inbound leads, MQLs (marketing-qualified), SQLs (sales-qualified, accepted by sales), Opportunities, Closed-won. The math is identical to a product funnel. The B2B-specific number to watch is the MQL → SQL conversion (marketing-to-sales handoff) — that's where most sales orgs lose users to definitional ambiguity, not real disqualification. The calculator surfaces it as a leak so you can have the conversation.
- How do I track stages over time without a tool like Mixpanel?
- For most teams under 1,000 users/month, a Google Analytics 4 funnel exploration plus a weekly tab in a spreadsheet is enough. The action: every Friday, log in, read the count at each stage for the last 7 days, paste into a spreadsheet column. After 8–12 weeks, you have a trendline. This calculator is the analysis layer on top — paste this week's numbers, see the conversion. Paid funnel tools (Mixpanel, Amplitude, Heap) become worth it when you need to slice by cohort or run experiments without engineering time.
- Does the order of stages matter?
- Yes — funnel stages are ordered by user journey, not alphabetically. Stage 1 must be upstream of stage 2 in real time. If your funnel is Visit → Signup → Purchase, and a user can purchase without signing up, your funnel isn't actually linear and the conversion math is misleading. Either restructure into multiple funnels (one per path) or define stage 1 as 'reached the product page' (the actual common upstream event for both paths). The calculator assumes linear, ordered stages.
- What's the difference between conversion rate and retention rate?
- Conversion measures a one-time action: did the user advance from stage A to stage B at any point. Retention measures repeat behavior: did the user come back N days later. Both can be modeled as funnels but mean different things. A retention funnel might be Day 1, Day 7, Day 30 — the 'stages' are time windows, not events, and the count at each is the number of users still active. You can use this calculator for retention curves; just label your stages 'Day 1', 'Day 7', 'Day 30' and enter user counts at each.
- Why does the simple formula work — can't a leak in step 2 hide a problem in step 5?
- It can, and that's a real interpretive risk. The headline overall conversion compresses every stage's performance into one number. A 10% overall conversion could be 100% × 10% (a single bad step) or it could be 50% × 50% × 40% (consistently mediocre across multiple steps). The calculator shows you both — the overall number and the per-step rates — exactly so you don't fall for the average. Always look at the per-step breakdown before claiming you 'know' what's wrong.
