Conversion Rate Calculator
One conversion rate tells you almost nothing about where you are losing people. This works out the headline number and the four step rates underneath it, so the biggest drop is visible — and prices what a ten per cent improvement would be worth on your own traffic.
INPUTS
Six
RUNS IN
Your browser
EMAIL REQUIRED
No
FORMULA
Printed below
The average is the number that hid the problem. The step rates are the number that finds it.
- Calculator
Your conversion rate, step by step
Pull six numbers from your analytics for the same date range — a full month is steadier than a week. Everything recalculates as you type, and nothing leaves your browser.
Conversion rate = orders ÷ sessions × 100. Each step rate divides the sessions that reached that step by the sessions that reached the one before it, so the four multiply back to the headline rate. The lift figure is orders × 0.1 × average order value — a ten per cent relative improvement, not ten percentage points.
- Reading it
The average rate is the one that hides the problem
A single percentage tells you that something is wrong. The four step rates tell you where.
A store at 1.8% looks ordinary next to the usual benchmarks, and that is exactly why the number is dangerous: it averages four separate behaviours into one figure and none of them are visible. Half of all sessions never reach a product page at all — that is a navigation and landing-page problem, and no amount of checkout work will touch it.
Read the four tiles in order and find the one that is furthest below where it should be. Product view rates below 40% usually mean traffic is landing somewhere it cannot act. Add-to-cart under 10% points at price, photography or stock. A checkout start rate under half says the cart page is doing the rejecting; a completion rate under 40% says the checkout itself is. Fix the worst step first, then re-measure.
- Use one date range of at least four weeks for all six inputs
- Segment by device before you act — mobile and desktop rarely leak at the same step
- Compare each step against itself last quarter, not against a published benchmark
- Getting it wrong
Four ways a conversion rate misleads
Each of these produces a number that is technically correct and practically useless.
Mixing the denominators
Orders divided by users is not orders divided by sessions, and the two can differ by a third. Pick one denominator, write it down, and never compare a rate built on sessions with one built on users.
Reading one blended number
Mobile usually converts at half the desktop rate. A blended figure moves whenever the traffic mix moves, so a rate can fall in a week where nothing on the site changed at all.
Chasing a published benchmark
Benchmarks average across categories, price points and business models that have nothing to do with yours. Your own rate last quarter, by device and by channel, is the only comparison that means anything.
Calling a two-day test a result
A rate calculated on a few hundred sessions moves on noise. Give any change at least two full weeks, and check the step rate you intended to move rather than only the headline.
Our conversion rate optimization guide covers research, prioritisation and testing in order — this calculator is the measurement step that comes before all of it. If you want the same three terms rolled into a revenue number, the eCommerce revenue calculator multiplies them out and projects them forward.
- Questions
Conversion rate FAQs
The questions this calculator usually raises.
What is a good eCommerce conversion rate?
Most eCommerce stores sit between 1% and 3%, but the range inside any single category is far wider than the gap between categories. High-consideration goods with a long research cycle convert below 1% and are perfectly healthy; low-price repeat-purchase categories can pass 5%. The useful comparison is your own rate, by device and by channel, against the same period last quarter — a benchmark cannot tell you whether 1.8% is good for a store selling what you sell at the price you sell it.
Should I divide by sessions or by users?
Sessions is the default in most analytics platforms and the convention this calculator follows. Dividing by users produces a higher number because one person may visit several times before buying, and neither is wrong — but mixing them is. Record which denominator you use, keep it constant, and never compare a session-based rate with a user-based one.
Why do my four step rates not multiply back exactly?
They will if all five session counts come from the same date range and the same segment. Small gaps usually mean one figure was pulled from a different report, a different attribution window, or includes sessions the others exclude — bot traffic and internal IPs are the common culprits. Rebuild all six numbers in one report before acting on them.
Is a 10% lift realistic?
As a relative improvement over several months, yes, and it is a deliberately modest target — a store at 1.8% moving to 1.98%. Programmes that fix a genuinely broken step often do considerably better in the first quarter and then slow. The tile exists to make the revenue case concrete before any work is commissioned.
Should I measure conversion rate by channel?
Yes, and it is usually the most revealing cut after device. Paid search, organic, email and social arrive with different intent, so a blended rate moves whenever the mix moves even if every individual channel is flat. Run the calculator once per channel and the picture changes quickly.
Does this handle a subscription or multi-step funnel?
The five-step structure follows a standard retail funnel. If yours differs — a quote request, a trial, a subscription sign-up — the arithmetic still holds: put the sessions reaching each stage in order and read the drop between them. Only the labels change.
- Next
Related tools and reading
CALCULATOR
Goes deeper on the two weakest steps here — what abandoned carts and checkouts are costing you each month.
GUIDE
Research, hypotheses, prioritisation and testing in order, once you know which step is leaking.
FREE AUDIT
Five working days across six disciplines, including a session-recording pass over whichever step your numbers point at.
Know which step is leaking but not why?
That is the harder half. The free audit watches real sessions through the step your numbers point at and comes back with the specific reasons people are stopping.