Definition (canonical)
Conversion rate is the share of people who take the desired action: `conversions ÷ visitors`. The "desired action" defines everything — purchase, signup, quote request — and a business runs on a CHAIN of conversions, not one: visit→cart→checkout→payment each has its own rate and its own leak.
Benchmarks with the necessary asterisks
E-commerce sitewide typically converts 1–3%; SaaS visitor→trial 2–5% and trial→paid 10–25%; lead-gen landing pages 5–15%. The asterisks matter more than the numbers: traffic source moves rates several-fold (brand search converts many times better than cold social), device gaps persist (mobile often half of desktop), and price/consideration length set gravity. Practical conclusion: benchmark against YOUR history segmented by source and device — a blended rate is a number that hides all decisions. Conversion also trades off against traffic quality: doubling ad reach into colder audiences "worsens" conversion while possibly growing profit — read it with CAC, never alone.
The CRO loop — research first, tests second
Sustainable optimization is a loop: research (session recordings, funnel analytics, user interviews, exit surveys — find WHERE and WHY people drop) → hypothesis ("checkout asks for account creation; guest checkout will lift completion") → test (A/B with pre-set sample and duration) → learn and repeat. Teams that skip research and A/B-test button colors generate noise with confidence intervals. The heavy hitters recur across industries: page speed, removing form fields, guest checkout, visible trust signals (reviews, security, verifiable company data), clarity of the value proposition above the fold, and price presentation.
Faked improvement — the classics
Rate up, business down: filtering out "bad" traffic improves the ratio while shrinking volume; adding friction pre-purchase (qualifying steps) boosts measured conversion of survivors; discounts convert today and train customers to wait. The honest scoreboard pairs conversion with absolute conversions, margin per conversion and downstream funnel quality — optimize the system, not the KPI.
FAQ
How long should an A/B test run? Until pre-calculated sample size AND at least full weekly cycles — stopping at first significance is how false winners ship.
Micro vs macro conversions? Macro = the money action; micro = steps toward it (add-to-cart, email capture). Micro diagnoses; macro decides.
Is personalization worth it? After the basics — a slow site with a confusing offer personalizes its confusion.
Related: E-commerce · Sales Funnel · CAC · Lead Reviewed: 2026-07-15 · Sources: CRO practice