Before you can optimise anything, you need to know how to calculate your current conversion rate - and what that number actually means in context. A 3% rate is outstanding for a high-consideration B2B purchase and poor for a simple newsletter sign-up. Context is everything.
The formula
Conversion Rate Formula
Conversion Rate = (Conversions ÷ Total Visitors) × 100
Express the result as a percentage. Use unique visitors, not sessions, for the denominator.
Worked example: e-commerce product page
Your product page receives 8,400 unique visitors in one month. Of those, 168 complete a purchase. Conversion rate: 168 ÷ 8,400 × 100 = 2.0%. Monthly revenue at a $110 average order value: $18,480.
Worked example: you run an A/B test adding a sizing confidence tool and a visible returns policy. Four weeks later, purchases rise to 235. New rate: 2.8%. That 0.8 percentage point improvement is a 40% lift in revenue - $7,370 more per month from the same number of visitors.
Worked example: SaaS free trial
Worked example: your pricing page receives 3,200 visitors monthly. 96 start a free trial. Rate: 3.0%. You rewrite the CTA copy from 'Start your free trial' to 'Start my free trial' and add a social proof strip. Trials rise to 128. New rate: 4.0%. Customer acquisition cost drops by 25% from the same ad spend - no traffic increase required.
The multiplier effect
Small percentage point improvements create disproportionate revenue gains because they multiply against your entire traffic volume. A 1-point lift is worth more to a site with 100,000 visitors than a 5-point lift is to a site with 5,000.
What counts as a conversion?
A conversion is any action that moves a visitor toward your business goal. It does not have to be a purchase. Define it based on what you're trying to achieve on that specific page:
- E-commerce: completing a purchase, adding to cart, initiating checkout
- Lead generation: submitting a form, calling a phone number, requesting a quote
- SaaS: starting a free trial, booking a demo, activating a key feature
- Media/publisher: subscribing to a newsletter, signing up for an event
- B2B: downloading a whitepaper, requesting pricing, booking a consultation
Conversion rate benchmarks by industry
Use these as directional guides, not absolute targets. Your real benchmark is your own historical rate. Beat that consistently, and industry averages become irrelevant.
Crow's benchmark methodology
Use industry ranges only when the publisher names its sample, date range, conversion definition, and segmentation. Otherwise, compare against your own historical rate and report the traffic source and device alongside it. This keeps a directional comparison from becoming a false target.
Unbounce's 2024 Conversion Benchmark Report meets the first part of that test: it draws on more than 41,000 landing pages, 464 million unique visitors, and 57 million conversions from the preceding year. Its most useful lesson is not a single universal percentage. It is that conversion performance changes materially by industry, paid traffic source, and device.
Segment before you compare
In some industries covered by Unbounce, mobile generates up to seven times more visits than desktop while still converting at a lower rate. A blended site-wide average can therefore hide both the channel producing the volume and the device experience losing the opportunity.
Measurement guidance: Google Analytics: recommended events
External benchmark report: Unbounce Conversion Benchmark Report (2024)
Frequently asked questions
What's a good conversion rate?
It depends entirely on your industry, traffic source, device, and what you're asking visitors to do. A single-step email sign-up should convert far higher than a multi-step checkout. Use a large, transparent industry study as directional context, then judge improvement against your own like-for-like historical rate rather than treating one abstract percentage as universally good.
How is CRO different from SEO?
SEO brings people to your site. CRO converts them once they arrive. They complement each other: more traffic amplifies CRO wins; better conversion rates improve the return on SEO investment. Neither substitutes for the other.
How much traffic do I need to run A/B tests?
The traffic required for an A/B test depends on your baseline rate, minimum detectable effect, significance threshold, and number of variants. Calculate those inputs before launch; low-traffic sites should focus on qualitative research and larger, bolder changes rather than presenting an unqualified visitor count as a rule.