How to Calculate Your Conversion Rate (With Formula and Examples)

The formula for conversion rate is the easy part. Everyone publishes it, and it has not changed in twenty years. What trips people up is everything around it: which number goes on the bottom, what actually counts as a conversion, why Google Analytics 4 and Shopify give you two different answers, and (if you run a UK site) why your consent banner quietly changes the maths.

This guide covers the formula, three worked examples for different business types, the GA4 terminology change most articles have not caught up with, and the consent caveat that UK practitioners cannot ignore.

The conversion rate formula

Conversion rate is the share of visits or visitors who complete the action you care about:

Conversion rate = (Conversions / Total sessions or visitors) x 100

So if 50 people buy out of 5,000 visits, that is (50 / 5,000) x 100 = 1.0%.

That is the whole formula. The hard work is deciding what each part means, and the bottom number (the denominator) is where most reported figures go wrong.

The denominator problem: sessions, visitors, or a scoped group

There are three legitimate denominators, and each one gives you a different headline number from identical conversion data.

• Total sessions: every visit, including repeat visits by the same person.
• Unique visitors or users: deduplicated individuals.
• A scoped population: for example only visitors who reached the pricing page, or only sessions the platform classes as engaged.

Here is the example that makes the problem obvious. Suppose one person visits your site five times before buying.

• Session-based: 1 conversion / 5 sessions = 20%.
• Person-based: 1 person converted / 1 person = 100%.

Same human, same single purchase, and the headline rate moves by 5x depending only on the denominator you chose. Neither number is wrong. They answer different questions:

• A session-based rate tells you the likelihood of a conversion per visit.
• A visitor-based rate tells you the likelihood that a person eventually converts.

The cardinal sin is mixing them. If you compare your session-based rate against a benchmark, a previous period, or another tool that used unique visitors, the comparison is meaningless. Pick one denominator, hold it constant across every report, and state it whenever you quote a number. “2.5% conversion” tells the reader nothing; “2.5% of sessions” tells them what you measured.

GA4: it is now called “key event rate”

This is the single most out of date thing in most published guides. In March 2024, Google renamed GA4 “Conversions” to “Key events.” The metric names changed with it:

• Session Conversion Rate became Session key event rate.
• User Conversion Rate became User key event rate.

Google made the change to separate two vocabularies that used to clash. In GA4 the important action is a key event; in Google Ads the same action is a conversion. So a purchase shows up as a “key event” in your Analytics reports and as a “conversion” in Ads. If that distinction has ever confused you, that is why. Google explains the split on its own Conversions vs. key events page.

GA4 gives you two rates out of the box:

• Session key event rate = sessions with at least one key event / total sessions.
• User key event rate = users who triggered the key event at least once / total users.

By default GA4 counts key events once per event, which is Google’s recommended setting and lets you tell apart a session with several key events from one with a single key event. You can see the full definition on Google’s About key events page.

To find the rate in the interface: go to Reports, then Acquisition, then Traffic acquisition. Session key event rate is now shown there by default; if you do not see it, click the pencil or customise icon, open Metrics, and add Session key event rate.

Three worked examples

Always write the maths out inline so the denominator is visible. (conversions / denominator) x 100 = rate.

Ecommerce (macro conversion: a purchase)

50 orders over 5,000 sessions:

(50 / 5,000) x 100 = 1.0%

That sits below the rough 2 to 3% ecommerce average, which is a useful prompt to investigate. Now watch the denominator do its work with the same 50 orders:

• (50 / 2,000) x 100 = 2.5%
• (50 / 1,000) x 100 = 5.0%

Same numerator, three very different rates. This is exactly why you cannot read a benchmark without knowing its denominator.

Lead generation (macro conversion: an enquiry form)

30 enquiry-form submissions over 1,200 sessions:

(30 / 1,200) x 100 = 2.5%

B2B lead-gen sites commonly convert around 2 to 3% of visitors into enquiries, so this is a fairly typical result.

SaaS (macro conversion: a free-trial start)

38 trial sign-ups over 1,000 landing-page sessions:

(38 / 1,000) x 100 = 3.8%

SaaS rarely lives on one number, though. The trial start is the first conversion; trial to paid is a second, separate conversion rate. So SaaS teams track a chain (visitor to trial, then trial to paid) rather than a single figure. Collapsing that into one rate hides where customers actually drop off.

Micro and macro conversions

Before you can calculate anything, you have to define the event. There are two kinds.

• Macro conversion: the primary business goal. A purchase, a completed lead form, a quote request, a paid subscription.
• Micro conversion: a smaller step toward that goal. Add to cart, reaching checkout, a newsletter sign-up, watching a product video, creating an account, downloading a guide.

Each micro and macro action has its own conversion rate, so “your conversion rate” means nothing until you say which event is the numerator. Micro conversions happen far more often, which gives you more data and faster CRO iteration; macro conversions are rarer and harder to reach statistical significance on in an A/B test. A clean way to hold the two in your head: the micro conversions are your diagnostic funnel, and the macro conversion is the scoreboard.

Why your GA4, Shopify, and CRM numbers never match

Three different tools will report three different conversion rates for the same week, and that is normal. The usual reasons:

• Different denominators. GA4 may report on sessions while Shopify reports on a different visit definition entirely.
• Different attribution. Each platform decides on its own which visit gets credit for the order.
• Consent gaps. Analytics that depends on cookies only sees the visitors who agreed to be tracked. Your order platform sees everyone who paid. See the next section.

The fix is not to make them agree. It is to nominate one tool as your source of truth for the conversion rate, document its denominator, and use the others for cross-checks rather than for the headline figure.

The UK angle: consent changes your denominator

This is the part most guides skip, and for a UK site it is not optional. Under the Privacy and Electronic Communications Regulations (PECR) and UK GDPR, analytics cookies are not “strictly necessary.” They serve the site operator, not the visitor, so they are not exempt and they require prior, informed, opt-in consent before they fire. The Information Commissioner’s Office sets this out in its guidance on cookies and similar technologies.

Valid consent has to be freely given, specific, informed, and an unambiguous affirmative action. Pre-ticked boxes and “by continuing to browse you agree” banners do not count.

The consequence for your calculation is direct: every visitor who rejects or ignores the banner is missing from your analytics, so they are missing from both the numerator and the denominator. A GA4-measured conversion rate in the UK is computed on a consented subset of traffic, not on all of it, which means it can be skewed against the true picture. GA4’s consent mode can model some of the unconsented activity to fill the gap, so the number is an estimate rather than a hole, but it is still not a full census of your visitors. Treat your analytics-based rate as “rate among consenting, measurable visitors,” and reconcile against your order or CRM totals, which capture everyone who actually converted.

This matters more now than it used to. The ICO finalised updated guidance on storage and access technologies on 29 April 2026, following the Data (Use and Access) Act, and the Act raised the maximum PECR penalty to up to £17.5m or 4% of global turnover, well above the old £500,000 cap. Consent-gated analytics is the UK default, not a nice-to-have, so it pays to understand how it bends your measurement.

What is a good conversion rate?

Benchmarks are useful as a sense check and dangerous as a target, mostly because almost none of them state their denominator. Treat the following 2025 to 2026 figures as directional only.

• All-industry web average: roughly 2.9%, with most sites between 1 and 4%.
• Ecommerce: around 1.8 to 3% on average; 3 to 5% is considered good; top performers clear 5%.
• B2B lead generation: roughly 2 to 3% visitor-to-lead, with strong landing pages above 5%.
• SaaS: around 1.1% blended for B2B SaaS, but the median SaaS landing page sits near 3.8%. Dedicated lead-gen landing pages can average far higher because their traffic is pre-qualified, which flatters the number rather than reflecting a better site.

These come from secondary marketing sources and move year to year, and clean UK-specific tables are not separately published in a form worth quoting. The more reliable benchmark is your own trend line. A site moving from 1.8% to 2.2% on a stable denominator has improved; the same site sitting at 2.2% against a stranger’s “industry average” of 2.9% has learned almost nothing, because you do not know what that stranger measured.

Calculating by channel, landing page, or campaign

To break the rate down by segment, scope the denominator to match the numerator’s population. If you want the conversion rate for paid search, divide paid-search conversions by paid-search sessions, not by total site sessions. If you want a landing page’s rate, use that page’s entry sessions as the denominator. Mismatching the two (counting conversions from one segment over traffic from a wider group) is one of the most common reporting errors, and it always makes the rate look worse than it is.

If you are setting up structured experiments to lift these segment rates, our conversion rate optimisation guide covers the testing workflow, and our roundup of the best A/B testing tools for 2026 compares the platforms that handle the statistics for you.

Frequently asked questions

What actually counts as a conversion? Whatever you define as a meaningful business action: a purchase, a completed lead form, a phone call, a demo request, a sign-up, or a download. The maths is trivial; the discipline is choosing one event, tying it to a real business goal, and tracking it consistently. Until you name the event, “your conversion rate” has no fixed meaning.

Should I divide by sessions or unique visitors? Either, as long as you stay consistent. Sessions answer “how likely is a conversion per visit,” and unique visitors answer “how likely is a person to convert eventually.” The danger is mixing the two across reports or comparing your session-based rate to a visitor-based benchmark. Pick one, document it, and never switch mid-comparison.

Is conversion rate still called that in GA4? Not since March 2024. GA4 renamed conversions to “key events,” so you now look for session key event rate and user key event rate. The action is a “key event” in GA4 reports but still a “conversion” in Google Ads, which is why the two products can describe the same purchase with different words.

Why doesn’t my GA4 number match my Shopify or CRM figure? Three reasons usually stack up: the tools use different denominators, they attribute the conversion to different visits, and your analytics only counts visitors who accepted cookies while your order platform counts everyone who paid. Nominate one tool as the source of truth for the headline rate and use the others to cross-check.

Does the UK consent banner really affect my conversion rate? Yes. Under PECR and UK GDPR, analytics cookies need opt-in consent before they fire, so visitors who reject or ignore the banner never enter your analytics. Your GA4 rate is therefore measured on a consented subset of traffic, not on everyone. Consent mode can model some of the gap, but reconcile against order or CRM totals for the true conversion count.

What is a good conversion rate for my site? Most sites land between 1 and 4%, with an all-industry average near 2.9%, but those figures rarely state their denominator, so use them only as a rough sense check. Your own trend on a fixed denominator is a far more honest benchmark than someone else’s headline number.

If you are choosing testing software to act on these numbers, see our comparisons of Optimizely alternatives and VWO vs Optimizely.