Guide
07 November 2026

Conversion rate calculation: formula, GA4 and benchmarks

You calculate conversion rate by dividing the number of conversions by the number of visitors, then multiplying by 100. The formula is simple; the trap is elsewhere. In GA4, dividing by sessions or by users gives two different rates (2% versus 3.08% in our example): the gap depends on how many visits each user makes. The median e-commerce rate in France sits around 2% (Lysible barometer, based on Contentsquare), though that benchmark tells you very little about your own site. Separate your macro and micro conversions, wait for enough volume, and compare yourself to your own history first.

liste de données analytic GA4 pour suivre son site internet
The conversion rate is the percentage of a website's visitors who complete a desired action (purchase, sign-up, contact), calculated by dividing the number of conversions by the number of visitors over a given period.

In brief

16 min read Level: beginner Tools: GA4, Google Search Console, Google Sheets

  • The basic formula: conversion rate = (number of conversions ÷ number of visitors) × 100, over a fixed period.
  • In GA4, the denominator is not neutral: for the same 200 buyers, our example gives 2% on 10,000 sessions and 3.08% on 6,500 users.
  • The median e-commerce conversion rate in France sits around 2%, between 1.5% and 3% depending on the sector, and the device changes everything. Lysible barometer, based on Contentsquare
  • Separate macro conversions (purchase, quote) from micro conversions (newsletter sign-up, add to cart): mixing them artificially inflates the rate.
  • Rule of thumb: under 100 conversions over the period, the rate moves too much to decide on; wait for enough volume.
  • Always compare yourself to your own past before comparing to the sector: the benchmark points the way, your history decides.

Contents

A conversion rate read off a dashboard doesn't always reflect real performance. Often the number doesn't say what you think it does: wrong denominator, period too short, micro and macro conversions mixed together. Before you can tell whether your rate is good or bad, you need to know what it actually measures. That's what this guide covers, from the by-hand calculation to the industry benchmarks.

Why the same site can show two different conversion rates

The same site can show two very different rates depending on whether you divide by sessions or by unique users: in the example below, the gap tops 50%. The formula doesn't lie; it's the denominator you pick badly.

Screen showing two dashboards side by side comparing conversion rate by sessions and by users
Two denominators, two numbers, and yet the same sales.

The calculation looks trivial. You take your conversions, divide by your visitors, multiply by one hundred. Except that "visitors" is not a neutral word inside a measurement tool. The same person can come back three times before buying. Whether you count them as three sessions or as a single user, the denominator changes completely.

Take a concrete case. A store gets 10,000 sessions in a month, generated by 6,500 unique users, and records 200 orders, placed by 200 different customers, one per visit. Calculated on sessions, the rate is 2%. Calculated on users, it climbs to 3.08%. Same month, same sales, and yet a gap of more than 50% between the two figures. Whoever steers by the first doesn't decide the same way as whoever steers by the second.

What is the formula for finding the conversion rate?

The formula is: conversion rate = (number of conversions ÷ number of visitors) × 100. You get a percentage over a fixed period. The difficulty never comes from the operation itself, but from what you put in each slot: which action counts as a conversion goal, and who counts as a visitor. Set that double definition before you even open your measurement tool.

Sessions or users: which denominator should you choose?

Choose sessions for a quick purchase or to optimize a page or a funnel, and users when the customer comes back several times before converting, which is the norm in B2B or for a quote. The official Google Analytics 4 documentation details these two metrics: GA4 now calls them "session key event rate" and "user key event rate" (formerly session conversion rate and user conversion rate). In both cases, GA4 counts the sessions or users that triggered the purchase at least once, not the number of orders: a customer who orders twice counts only once in the numerator of the per-user rate.

What matters most is consistency. Set one denominator and never change it again. The day you switch from sessions to users without noting it, you think you've improved (by more than 50% in our example). Yet you've changed nothing other than your calculation.

For Google Ads campaigns, a third denominator applies: clicks. It measures the return of a paid campaign on a cost-per-acquisition basis. It's not meant to be compared to the other two. Each answers a different question; mixing them produces false conclusions. So decide from the start, and write that choice down somewhere.

The exact formula, step by step, with a worked example

Applying the formula cleanly comes down to four decisions: the period, the action that counts, the denominator, then the calculation. The common mistake isn't in the division; it's in the order and the rigor of those decisions.

Let's take a small business selling sports gear online, a catalog of about 300 items, with strong spring seasonality. For the month of April, GA4 reports 8,400 sessions. The store wants to know its purchase conversion rate.

First decision: the period. The whole of April, from the 1st to the 30th, never a single week. Second: the action. Here, a completed order, not an add to cart. Third: the denominator. We keep sessions, because we're optimizing the funnel. Fourth: the calculation. 168 orders ÷ 8,400 sessions × 100 = 2%.

100

Below 100 conversions over the period, your rate stays fragile: a handful of sales more or less shifts it by several tenths of a point.

Estimate: rule of thumb for monthly tracking, not a universal statistical threshold

How do you work out a conversion calculation?

Take your conversions for the period, divide by your visitors for the same period, multiply by 100. Make sure the two figures cover exactly the same dates and the same scope, meaning the same traffic source and the same device if you segment. A clean calculation starts with two comparable figures at the numerator and the denominator.

Over what period should you calculate for a reliable figure?

Calculate over a period that adds up to at least 100 conversions, ideally a full month to smooth out the quiet days. A week rarely suffices for a small business: a weekend, a public holiday or a stockout skews the reading. The calendar month has another advantage, it compares easily year over year. Strong seasonality? Never compare April to January; compare April to the previous April. This rule keeps you from mistaking a real trend for a mere calendar effect.

How to calculate a store's conversion rate in GA4

In GA4, a store's conversion rate is read in the "Traffic acquisition" report, by choosing purchase in the "Session key event rate" column; the figure shown by default mixes several key events. So before reading it, you need to know what you've defined as a key event.

GA4 has changed its vocabulary: the interface now talks about "key events" rather than "conversions". Every event flagged as key enters the numerator of the rate. If you've flagged purchase, generate_lead and newsletter_signup all at once, your displayed rate counts any session where one of the three occurred. It no longer represents your purchase rate, but a composite rate that no longer means much.

So for a store, isolate the purchase event: GA4 then recalculates the rate on that single event, source by source. That's the figure to read, not the overall rate on the reports home page. Marking the purchase usually goes through Google Tag Manager, the tag manager that fires your events at the right moment. To go further in setting up sales tracking, our guide to configuring GA4 on an e-commerce store structures events cleanly before you calculate anything at all.

Which calculation should you use for an online store?

Divide your completed orders by your sessions over the same period, isolating the purchase event in GA4. Don't use the composite rate shown by default if you have several active key events. Also check that your purchase tag fires only once per order: a double fire inflates the numerator. A check in Google Tag Manager's preview mode is enough to spot that duplicate.

Where do you read the conversion rate in GA4 reports?

The right place is the "Traffic acquisition" report, with purchase selected in the rate column: it isolates the purchase, which generates revenue, and gives its rate source by source. The home-page tile aggregates all key events without distinguishing the purchase from the rest: low reliability for a store. The "Purchase journey" report, under Monetization, complements it by showing at which step sessions drop off. Free-form exploration, finally, is as reliable as Traffic acquisition and more flexible: there you choose the metric yourself, session rate or per-user rate, and segment freely.

And what about your own site?

The analysis measures speed, technical SEO and Core Web Vitals, then gives you 3 concrete recommendations. Free, no account needed.

Analyze my site for free

Macro conversions and micro conversions: don't mix everything

A macro conversion is the action that keeps your business alive; a micro conversion is an intermediate signal. Adding them up in a single rate artificially inflates the number and blinds you to what matters.

It sometimes happens that a rate reported at 8% or 9% drops to 2% as soon as you keep only the sales. The difference is the micro conversions counted as wins. An add to cart is not a sale. A newsletter sign-up generates no revenue. These signals have value, yes, but not the same value, and above all not in the same indicator.

The healthy reflex: one rate per type of action, never a catch-all rate. You track your purchase rate, your add-to-cart rate and your sign-up rate separately. You then see where it gets stuck. If 30% of visitors add to cart but 2% buy, your problem isn't at the top of the purchase funnel, it's at checkout. Data from the Baymard Institute puts the average cart abandonment rate at around 70%, which places checkout at the heart of diagnostics. A single composite rate would have hidden this signal from you.

Should add-to-cart events count as conversions?

Count add-to-cart events as micro conversions to track separately, never mixed into the purchase rate. Add to cart is an excellent indicator of interest and a useful measurement point for spotting funnel drop-offs. But it generates no revenue until the order is completed. Keep it in your tracking, on its own line, alongside your macro conversion.

Macro conversions: three worked examples

One action per rate, never added together

168 orders ÷ 8,400 sessions = 2%

Completed purchase

42 quotes ÷ 3,500 sessions = 1.2%

Quote request

25 bookings ÷ 2,000 sessions = 1.25%

Appointment booking or call

Micro conversions: track each one separately

Signals of interest, not sales

2,520 adds ÷ 8,400 sessions = 30%

Add to cart

420 sign-ups ÷ 8,400 sessions = 5%

Newsletter sign-up

336 accounts ÷ 8,400 sessions = 4%

Account creation

168 downloads ÷ 8,400 sessions = 2%

Guide download

5,040 clicks ÷ 8,400 sessions = 60%

Click on "view product"

The calculation errors that distort your conversion rate

Most of the gaps between a GA4 rate and a store's real sales come from measurement bias, not a genuine conversion problem. Before reworking your pages, check that your figure is clean.

Diagram The biases that distort your GA4 rate Most of the gaps between a GA4 rate and real sales come from measurement bias, not a genuine conversion problem.
  1. Cookie consent In e-commerce, nearly one visitor in two refuses tracking and disappears from GA4.
  2. Mixed sources Back-office orders divided by truncated GA4 sessions: the rate looks better than it is.
  3. Unfiltered traffic Your visits, your agency's and spam inflate the denominator without converting.
  4. Double firing A tag counted twice inflates the numerator this time.

The first bias is cookie consent. Since the consent banner became mandatory, a share of visitors refuse tracking. Consent Mode v2 passes this choice to Google's tags: in basic mode, these visitors no longer appear in GA4; in advanced mode, Google can model some of them if the volume is sufficient. The CNIL strictly regulates this collection of consent, and the refusal rate varies widely from site to site. In e-commerce, nearly one visitor in two refuses. In GA4 alone, the visits and purchases of these visitors disappear together: the rate stays readable, but the volumes are underestimated. The trap appears when you divide your back-office orders, all counted, by your GA4 sessions, stripped of the refusals: your rate then looks better than it is. To grasp the scale of the phenomenon for your own business, the cookie consent rate by sector gives concrete benchmarks.

The second bias is unfiltered traffic: your own visits, your agency's and spam inflate the denominator without ever converting. GA4 already excludes known bots automatically; for internal traffic, it offers a filter, but you still have to turn it on. The third bias is sneakier: double tag firing, which counts two purchases for a single order and inflates the numerator this time.

~50%

On many French e-commerce sites, nearly one visitor in two refuses measurement cookies, creating a blind spot in GA4 that the conversion rate doesn't reflect.

Source: Lysible, cookie consent benchmark by sector (Didomi data)

This reconciliation between your real sales and what GA4 shows is done by hand, bias by bias, once a month. The goal isn't to reach a zero gap, but to know yours and keep it stable.

Why doesn't my GA4 conversion rate match my real sales?

Your GA4 rate diverges from your real sales because of cookie consent, unfiltered traffic and possible tag duplicates. GA4 only sees the visitors who accepted measurement, whereas your back office records every order, consented or not. The gap is normal and can be large: it depends mainly on your cookie acceptance rate. The good habit: compare each month the number of orders in your Shopify, WooCommerce or PrestaShop store with the number reported by GA4, to know your loss rate and keep it stable.

Is your conversion rate good? Benchmarks by sector

A decent e-commerce conversion rate sits between 1.5% and 3%, but that benchmark tells you almost nothing until you've segmented it by sector and by device.

Hand near a printed page of sector conversion rate benchmarks lying next to a keyboard
A page of sector benchmarks places your rate, without saying whether it's improving.

Reference figures vary across sources, and that's normal: each measures a different sample. Our e-commerce conversion rate barometer for France puts the median around 2%, based on Contentsquare, and the top 20% of stores above 3.2%. Shopify, for its part, considers a 3% rate satisfactory for an e-commerce site. Gaps between sectors remain marked: from 0.8% to 1.5% in furniture to 2.8% to 3.5% in food. In B2B, where the cycle is long, a 1% rate can be excellent. In fast-moving consumer goods, 3% can be disappointing.

The most underestimated gap is the one tied to device. Mobile almost always converts worse than desktop, often by half, even though it holds most of the traffic. According to StatCounter, mobile accounts for more than half of web visits worldwide. An overall rate of 2% can hide 3% on desktop and 1.2% on mobile: your priority task then jumps out at you. Reading an overall rate without breaking it down by device means steering without looking at the segment that holds most of the visits.

What is a good conversion rate in e-commerce?

A good e-commerce conversion rate generally sits between 1.5% and 3%, but that benchmark depends on the sector, the average order value and the device. Low-priced, impulse-bought products convert higher; expensive or technical products, lower. A high average order value justifies a lower rate, because each sale brings in more. Judge your rate against your profitability, not an average figure read on a blog.

Why comparing your rate to the sector can be misleading

Comparing your rate to the sector is misleading because samples, conversion definitions and device mixes differ from one study to another. A sector average aggregates thousands of sites with business models very different from yours. The only benchmark that really settles it is your own history: your rate this month against the same month last year, with an identical measurement scope. It's the same principle as for analyzing the bounce rate and session duration by sector: your own trend matters more than the sector average.

E-commerce conversion rate by sector and by device

Segment Observed conversion rate
Food / grocery 2.8% to 3.5%
Beauty and cosmetics 2.5% to 3.1%
Health and parapharmacy 2.3% to 3%
Fashion and apparel 1.8% to 2.5%
Tech and electronics 1.4% to 2.2%
Furniture and home decor 0.8% to 1.5%
Services and subscriptions 2% to 4%
Desktop, all sectors ~3%
Mobile, all sectors ~1.5%
Source: Lysible e-commerce conversion rate barometer for France (Contentsquare and Statista data, field observations).

Turning a one-off calculation into reliable monthly tracking

A conversion rate calculated once is useless; its value comes from the trend, month after month, at a constant scope.

Before touching a page, wait until you have three months of comparable rates in front of you. A good month can come from a promo, a bad one from a payment outage. Without history, you mistake noise for signal. It's this discipline of regular tracking that separates a solid diagnosis from an impression.

Five moves turn a one-off calculation into real steering.

  1. Set the denominator once and for all Sessions or users, choose and never change, otherwise your comparisons over time become false.
  2. Isolate the macro conversion Track your purchase or quote rate separately, apart from micro conversions, on its own line.
  3. Recalculate each calendar month Same method, same relative dates, to get figures you can stack month over month.
  4. Segment by device and by source An overall rate hides the real problems; desktop, mobile and acquisition channel tell different stories.
  5. Compare to your own history Your April rate against the previous April's, at an identical scope, before any sector average.

Sector averages aggregate realities too disparate to guide a precise decision. Also remember to cross these data with Google Search Console, which tells you which queries bring the traffic behind each rate.

How often should you recalculate your conversion rate?

Recalculate your conversion rate once per calendar month, over a period that adds up to at least 100 conversions. A monthly rhythm smooths out weekly swings and compares easily year over year. If your volume is low, under 100 conversions a month, switch to a quarter to keep a reliable figure. Checking your rate every day only produces noise, and you'd be reacting to meaningless swings. This logic of regular steering also applies to increasing the sales of an e-commerce site, where the conversion rate is the most profitable lever before any traffic spend.

Setting a clean figure and taking action

At this point, you know how to calculate your rate cleanly, choose your denominator, separate your conversions and clean up your biases. What's left is the hard part: doing it every month, without error, and turning each gap into concrete action on the right page. That's where most owners drop off, for lack of time. Lysible reads your Google data (Analytics, Search Console, Ads) and your pages at the same time, then tells you what to fix, how long it takes and for what gain. Reconciling with your orders still has to be done by hand: Lysible doesn't read your back office. Expect €89 to €129 excl. VAT per month, with a 30-day trial to judge on your own figures.

Frequently asked questions

What is the difference between conversion rate and transformation rate?

In everyday usage, none: both expressions refer to the percentage of visitors who complete the target action. "Transformation rate" is used mainly in e-commerce and retail, for the share of visitors who buy. "Conversion rate" is broader and applies to any action: purchase, quote, sign-up. What matters is to specify, each time you cite the figure, the action measured and the denominator used, sessions or users.

Is a 1% conversion rate necessarily bad?

No, a 1% rate can be excellent depending on the context. In B2B with a long cycle and a high basket, 1% of qualified quote requests beats 5% of clicks that go nowhere. For an expensive or technical product, a visitor who converts brings in a lot, which justifies a lower rate. Judge your rate against your real profitability and your average order value, never against an absolute threshold read on a general-interest blog.

How do you calculate a conversion rate in Google Sheets?

Create three columns: the month, the number of conversions (orders or quote requests), the number of sessions or users over the same period. In the fourth column, the formula =B2/C2 gives the rate; apply the percentage format to display it. Add a row per month without ever changing the denominator, then a line chart to track the trend. Enter the figures on the same date each month: regularity matters more than the tool.

How do you improve a conversion rate that's too low?

Start by checking that your figure is clean before touching your pages: consent bias, internal traffic, tag duplicates. Once the rate is reliable, segment by device to find where it gets stuck, often on mobile. Then look at your funnel: if add to cart is strong but purchase is weak, the problem is at checkout. Work on page speed, the clarity of the offer and the simplicity of the form before buying more traffic.

Is the Google Ads conversion rate the same as the GA4 one?

No, they almost always differ, and that's normal. Google Ads calculates its rate on clicks from its campaigns, with its own attribution model and its own conversion window. GA4 calculates on sessions or users, with potentially different attribution and event definitions. Don't try to make them match to the exact figure: use Ads to steer your campaigns on a cost-per-acquisition basis, and GA4 for your overall multi-channel rate.

Should you separate mobile and desktop conversion rates?

Yes, it's one of the most profitable segmentations. Mobile often converts half as well as desktop while holding most of the traffic. An overall rate masks this gap and stops you from prioritizing. By breaking it down by device, you immediately see whether your task is the mobile experience, such as a form that's too long or a slow page, or a cross-cutting problem. In GA4, simply add the "Device category" dimension to your conversion exploration.

Can you calculate a conversion rate without GA4?

Yes, provided you take orders and visits from the same tool, for example Matomo with its e-commerce tracking. If you divide your back-office orders by the visits from a measurement tool, that tool must count every visit, like a Matomo configured for consent-exempt audience measurement; otherwise the rate is overstated, from the same bias as in GA4. Some small businesses calculate their rate in a simple spreadsheet, entering their sales and traffic each month: the method matters less than the consistency of the two figures.

Conversion rate calculation: formula, GA4 and benchmarks

Isaac SIKORSKI

With Lysible, I want to give businesses back control of their online presence. A website you actually understand is one that brings in real opportunities.