Definition
Conversion rate is conversions divided by the opportunities to convert, expressed as a percentage. The numerator is whatever the business has defined as a conversion: a purchase, a lead form, a sign-up, a call, a key event in the analytics tool. The denominator is the unit of opportunity, and it varies by context. Ad platforms divide by clicks (or interactions). Web analytics divides by sessions or by users; Google Analytics reports both a session-based and a user-based key event rate, and they differ because one person can have many sessions. On-site experiments usually divide by users exposed to a variant.
A conversion rate without its denominator stated is not a metric, it is a rumour. The key events and conversions guide explains how the analytics definitions are configured and why platform and analytics figures rarely match.
Conversions ÷ Sessions (or Clicks)(%)
How to interpret it
Direction is "higher is better" for a fixed audience and a fixed definition of conversion. The qualification matters: a rate can rise because the site improved or because the traffic mix shifted towards people who were going to convert anyway, such as returning customers or brand searchers. Rates differ systematically by source (brand search converts far above prospecting display), by device (desktop usually above mobile for considered purchases), by landing page, by country and by new versus returning visitor. Any change in the blended rate should first be decomposed by those dimensions before anyone claims the site got better or worse.
The rate also interacts with the stages above it. Narrowing targeting raises conversion rate and lowers volume; broadening does the reverse. A higher CTRShare of impressions that result in a click. from a more enticing ad can lower conversion rate by attracting the curious. Conversion rate is therefore read alongside cost per acquisition and volume, never alone, and the CRO hub frames it as one outcome of the whole FunnelModel of the customer journey stages. rather than a property of a page.
What moves it
- Traffic quality and intent. The dominant factor; the same page converts brand traffic and cold traffic at very different rates.
- Message match. Whether the landing page delivers what the ad or listing promised.
- Page experience. Speed, clarity, form length, trust signals and payment options.
- Offer and price. Discounts and free delivery thresholds change the rate directly.
- Device and context. Mobile conversion depends on checkout design more than desktop does.
- Definition and tracking changes. A new key event, a consent banner or a broken tag shifts the rate with no change in behaviour.
- Seasonality, which moves intent long before it moves the site.
The last two are the ones to rule out first when a rate moves, because they change the number without changing anything a visitor experienced.
Reading it in context
| Observation | Likely cause | Confirm with |
|---|---|---|
| Blended CVR fell, every channel's CVR flat | Mix shift towards lower-intent traffic | Rate by source and campaign |
| CVR fell on one date, all channels | Tracking, consent or site release | Tag health, release log |
| CVR up, cost per acquisition up | Costlier clicks that convert better | Whether the CPA target still holds |
| Mobile CVR far below desktop | Checkout friction on small screens | Funnel step drop-off by device |
| Variant CVR higher in a test | Possibly real, possibly noise | Test significance and sample size |
A fall that was a mix shift
A retailer's site conversion rate drops from 3.2% to 2.6% in a month and the CRO team is asked to explain. By source, brand search converts at 8.0% both months, prospecting social at 1.1% both months, and email at 5.5% both months. What changed is the mix: a new prospecting campaign added 40,000 low-intent sessions, taking social from 20% to 38% of traffic. Nothing on the site got worse. Revenue rose because the campaign still produced 440 orders at an acceptable CPA. The conversion rate calculator gives the per-source rates in seconds.
Related metrics
CPACost of each conversion. is cost per click divided by conversion rate, so every movement in the rate shows up directly in acquisition cost, which is why conversion work is the cheapest way to lower CPA. CTRShare of impressions that result in a click. is the rate one step earlier in the chain, and the two often move in opposite directions when creative changes who clicks. The FunnelModel of the customer journey stages. view breaks a single rate into the step rates that explain it, from landing to basket to payment, and an A/B TestControlled comparison of two variants. is the only reliable way to attribute a change in rate to a change on the site rather than to traffic. The core metrics guide places conversion rate in the chain from impression to order. Reported together, these show whether a rate change was caused upstream, on the site, or by measurement.
App stores have their own version of this metric, and the app store listing conversion rate walk-through shows how to calculate it from page views and first-time installs.
Common mistakes
- Comparing rates with different denominators, such as a platform's click-based figure with analytics' session-based one, and then arguing about the gap.
- Benchmarking against industry averages that mix every source, device and conversion definition into one figure.
- Claiming a site improvement from a blended rate without checking whether the traffic mix moved.
- Optimising the rate by shrinking the audience, which raises the percentage and lowers the revenue.
- Reading a test result before it has the sample to support it, then shipping a change that did nothing.
- Changing the conversion definition mid-series without annotating the report, so the history becomes unreadable.
- Reporting one rate for the whole site when the pages, intents and devices behind it convert at rates that differ by an order of magnitude.
