Benchmarks
What is a good conversion rate for a WooCommerce store?
Everyone quotes a number and almost nobody says where it came from. We read the sources, and they disagree with each other by more than double. Here is why, what actually moves your rate, and the baseline you should be comparing against instead.
What this article covers
- The published averages disagree with each other by more than double. We went and read them. One credible source says 1.4%, another says 3.0%, and both are measuring honestly.
- Which means a single benchmark number cannot tell you whether you have a problem. Your industry, device mix and traffic source move it more than anything you do to the site.
- Your own baseline is the only useful comparison, and you can build one this afternoon from data you already have.
- Checkout completion rate is the number to actually watch. It strips out the browsing traffic that makes site-wide conversion rate so noisy.
The short answer, then the honest one
The short answer people want: somewhere between 1% and 3% is normal for most ecommerce stores, and above 3% is good.
The honest answer: that range is so wide it is nearly useless, and the sources it comes from disagree with each other by more than a factor of two. Before you use any benchmark to judge your store, it is worth knowing how soft the ground is.
Why the published numbers disagree so badly
We read the sources rather than repeating a number from a listicle. Here is the actual spread.
| Source | Says the average is | Measured across |
|---|---|---|
| Littledata | 1.4% | 2,800 Shopify stores, 2023 |
| Widely republished figure | 2.5% to 3.0% | Source rarely stated |
| Littledata, by device | Mobile 1.2%, desktop 1.9% | Same 2,800 stores |
| Widely republished, by device | Mobile 2.87%, desktop 3.09% | Source rarely stated |
Same metric, same year, more than double the difference. Neither is lying.
The gap is not dishonesty, it is definition. Four things move a published average enormously before any store optimisation enters the picture:
- Whose stores are in the sample. A dataset drawn from stores using a paid analytics tool skews toward larger, more established sellers. A dataset drawn from every store on a platform includes thousands of abandoned hobby shops converting at zero.
- What counts as a session. Bot filtering alone moves this by a wide margin. A store with heavy scraper traffic and no filtering will report a conversion rate roughly half its real one.
- Which industries are in the mix. Consumable goods with repeat buyers behave nothing like considered purchases. Blending them produces a number that describes neither.
- Traffic source. Branded search converts several times better than cold paid social. Two identical stores with different marketing mixes will report very different conversion rates.
The practical consequence. If you take a benchmark from an article, you cannot tell whether you are above or below it, because you do not know whether your store belongs in that sample. A number you cannot position yourself against is decoration. Treat every published average, including the ones in the table above, as a rough sense of scale rather than a target.
What moves the number more than your website does
Worth internalising before you spend a month redesigning something. In rough order of impact:
- What you sell. Consumables and low-consideration goods convert several times better than considered, high-value purchases. Nothing on your site closes that gap, and nothing should try to.
- Price point. Conversion rate and average order value trade off against each other almost mechanically. A store converting at 0.9% on a 400 order can be far healthier than one converting at 4% on a 12 order.
- Traffic source. Someone who searched your brand name is a different person from someone who tapped an ad in a feed. Blending them hides both.
- New versus returning. Returning customers convert several times better. A store with a large repeat base has a flattering blended number and may still be failing every first-time visitor.
- Device mix. Mobile converts lower everywhere. A store with 80% mobile traffic will report a lower blended rate than one at 50%, with no defect anywhere.
- Seasonality. The same store can move by half in either direction across a year.
Only after all of that does the actual quality of your product pages, cart and checkout show up. Which is not an argument that they do not matter. It is an argument that a blended site-wide number is a bad instrument for measuring them.
How to build your own baseline this afternoon
This is the part that replaces the benchmark, and you already have the data.
- Take twelve months, not one. A month is seasonality, not signal.
- Split it by device. Mobile and desktop separately, always. A blended number hides the device that is actually broken, and mobile is where most defects live.
- Split it by new and returning. Your first-time visitor conversion rate is the honest measure of whether the store works. Returning customers would buy through a worse one.
- Split your top three traffic sources. If paid social converts at a tenth of branded search, that is a media problem wearing a website costume.
- Write the six resulting numbers down. That is your baseline. It is worth more than any industry average because it is the only comparison that controls for what you sell, what you charge and who you buy traffic from.
Now you have a real question. Not "is 1.8% good", which is unanswerable, but "why does mobile convert at 40% of desktop for first-time visitors from search". That one has an answer, and you can go and find it.
The number to watch instead
Site-wide conversion rate mixes together everything from someone who bounced off the homepage to someone whose card was declined. It is a business metric, not a diagnostic one.
Checkout completion rate is the diagnostic. Of the people who started checkout, how many placed an order? That strips out all the browsing traffic and every upstream problem, and what is left is almost entirely under your control.
The scale it lives on is better documented, because the Baymard Institute has tracked it for fourteen years. Their documented average cart abandonment rate is 70.19%, across 50 studies. Of the shoppers who abandon, 42% say they were only browsing and never intended to buy, so that share was never yours to win.
What is left is the part worth measuring. Baymard's benchmarking of 344 ecommerce sites found that 65% have a mediocre or worse checkout experience, only 2% qualify as good, and none reached state of the art. Their estimate of the addressable gap is a 35% increase in conversion rate for the average large site, from an average of 32 distinct improvements.
Read that 35% carefully, because it is widely misquoted. It is an estimate of the total available from fixing everything, on a large site, measured against a rigorous 110-point standard. It is not what you get from installing something. Anyone quoting it as the result of a single change, us included, is misusing it.
What a realistic improvement actually looks like
Store owners consistently expect the wrong shape of result, which is why so many conclude that optimisation does not work.
- Single changes move conversion by fractions of a percentage point. Moving from 1.8% to 2.0% is a 11% relative improvement and a very good outcome for one change. It will look like nothing on a chart.
- The compounding is where the money is. Six changes each worth a few percent relative, applied to the same traffic, is a materially different business at the end of a year.
- Some of the gap is not design at all. Shoppers who do not trust the payment form leave without complaining, and that is 19% of ready-to-buy abandonments. See why shoppers do not trust your store with their card.
- Removing a defect beats adding an optimisation. A payment button that is untappable on some phones is not a conversion rate problem, it is a broken store, and fixing it produces a step change rather than a lift.
- Most published case studies are not replicable. A 30% lift headline usually comes from a small sample, a short test, or a store that started from a genuinely broken baseline.
And before you interpret any movement at all, work out whether you have enough orders for the movement to mean anything. Most small stores do not, and we wrote the arithmetic out in how many orders before your numbers mean anything.
So what should you actually do
- Stop comparing yourself to a published average. You cannot tell whether you belong in its sample.
- Build the six number baseline above, and re-read it every quarter.
- Watch checkout completion rate as your diagnostic, because it is the part you control.
- Find your worst gate before optimising anything. The method is in your traffic is fine, here is where the money leaks.
- Fix defects first, then optimise. A broken thing is not a low conversion rate, it is a broken thing.
See the step people actually leave at
OptiCheckout reports drop-off per checkout step rather than one abandonment number, so the diagnostic above stops being a spreadsheet exercise. Every template is rendered from the real plugin, so you can walk one on your own phone first.
Read next
DiagnosisYour traffic is fine. Here is where the money leaks.
The four gates between a visitor and an order, how to measure each one, and what the pattern of your drop-offs tells you about which problem you actually have.
12 min read ConversionHow to reduce cart abandonment in WooCommerce
The ranked reasons people abandon, what each one is worth, and the two highest value fixes, neither of which needs a plugin.
13 min read MeasurementHow many orders before your numbers mean anything?
The arithmetic that ends most small-store A/B tests before they start, and what to do instead when you are too small to test.
10 min read