Is my conversion rate bad? Real 2026 benchmarks by sector

A conversion rate on its own cannot tell you whether your site is broken, because one month on one platform produced rates from 0.55% to 5.23% depending only on what the shop sold.

You looked it up at midnight. You did not like the answer. A blog told you the average is 2 to 3 percent, you are sitting at 0.8, and now you are quietly certain the site is a disaster. Maybe it is. But the number you measured yourself against was built from businesses that sell nothing like what you sell.

Is my conversion rate bad?

There is no single answer, because the rate depends on what you sell and who arrives. In July 2026 IRP Commerce measured ten UK and Irish ecommerce sectors on one platform, using one definition. Rates ran from 0.55% to 5.23%. Same month, same maths, nine times the spread.

That spread is not about good shops and bad shops. It is about baskets. Arts and crafts sells a bag of yarn to someone who already wants yarn. Baby and child sells a pram, and nobody buys a pram on the first visit. Not diagnosed the rest yet? Run the traffic but no sales diagnostic first.

Is 0.8% a bad conversion rate?

Not automatically. 0.8% sits below the 2.26% IRP measured across all its markets in July 2026, and above the 0.55% it measured in baby and child that same month. Whether 0.8% is bad depends on your category, your average order value and how much of your traffic is mobile.

In that same table, baby and child had the worst rate of the ten sectors and an average order of £737.47, against a market average of £123.37. Worst rate. Six times the basket.

How do you tell which one you have?

Stop measuring your site against an industry and start measuring it against itself. Open GA4, go to Reports, then Tech, then Device category, and read session key event rate for mobile and for desktop. Two numbers, under two minutes. The gap between them belongs to you, not to your category.

Now the threshold. Across the 2,800 Shopify sites in Littledata's benchmark, mobile converted at 1.2% and desktop at 1.9%. A mobile rate around two thirds of desktop is ordinary. If yours is under half, that gap is a site problem, and the panels disagree on what normal looks like: see mobile conversion rate lower than desktop.

Second check, same two minutes. Traffic acquisition, session key event rate by channel. If one channel drags the blended number down, that is a targeting question rather than a page question, and is it your traffic or your website settles it.

Rather not do it by hand? Run a free Revslip audit and see the gap priced.

Why does this cost you money?

Because a borrowed benchmark sends you to the wrong work. Chase a 3% figure built from a different category and you rebuild pages that were already fine. Accept 0.8% because a blog called it normal and you leave a broken mobile checkout running for a year.

Your blended rate is a weighted average of your traffic mix, and IRP recorded mobile at 62.3% of July 2026 sales. For most shops the blended number is the mobile number wearing a disguise.

Put real figures through the leak formula, working from how to calculate conversion loss, traffic × CVR gap × AOV × mobile weight. Say 18,400 sessions a month, 62% mobile, mobile at 0.61% against desktop at 1.84%. Closing mobile to two thirds of desktop is 11,408 sessions times a 0.62 point gap, so 71 more orders. At £82.08 that is £5,828 a month. The size of the prize, not a forecast. But it is a number you can hold, and "below average" is not.

What is a good conversion rate by industry?

Here is one month of measured trading data instead of a recycled average. IRP publishes conversion rate by sector from its own platform, calculated as transactions divided by sessions. These are its ten UK and Irish markets for July 2026, ranked.

The sector with the worst conversion rate on that list is the one selling the most expensive baskets. Its rate is not a failure. It is a category.

Google agrees, quietly. Its Analytics benchmarking documentation publishes no cross-industry average at all, only the median plus the 25th and 75th percentile inside a peer group you pick.

Get your own numbers priced free, in about a minute.

How do you fix it?

The fix is not a page change. It is replacing one misleading number with four honest ones, then acting on the worst of them. Four steps, each with a way to check you did it right, and the whole sequence takes an afternoon.

1. Split before you judge. Device, then channel, then top landing page. Check: four numbers, not one.
2. Set your own baseline. Your best month in the last twelve. Check: your site already produced it, so you know it is reachable.
3. Price the gap. Run the leak formula on the worst segment. Check: you end up with a monthly money figure, not a percentage.
4. Fix that segment only. Check: it moves and the others hold still. That rules out the season.

How will you know it worked?

Watch the segment, not the blended rate. If mobile was 0.61% and desktop 1.84%, success is the ratio between them closing toward two thirds. Give it four weeks and at least 400 mobile sessions, and compare against the same weeks last year rather than last month.

When should you ignore this?

Two situations where this post is the wrong advice. If you sell one expensive thing to few people, revenue per session is your real metric and your rate will look bad forever, correctly. And under a thousand sessions a month, three extra orders move the rate a full point, so it is too noisy to act on.

Worth saying about us. Revslip measures the site and prices the gap, but it cannot tell you whether your prices are wrong for your market, and price is a big lever here. Our own evidence has the same shape of hole. Across the 134 businesses Revslip has audited, the rate an owner quotes is almost always the blended figure from the GA4 home card. Those 134 are not a random sample, though. They are sites whose owners already suspected something was wrong.

Common questions

Three questions that come up every time this number gets discussed, answered from the same data as above. If yours is not here, split your own traffic and compare the segments.

What is the average ecommerce conversion rate in 2026?

IRP Commerce measured 2.26% across all its markets in July 2026, up from 1.94% a year earlier. Treat that as directional. It covers UK and Irish B2C merchants on one platform using last-click attribution, and the spread underneath runs from 0.55% to 5.23%.

Why does Google Analytics not show me an average?

Because Google decided a single cross-industry figure would mislead. Analytics benchmarking reports the median plus the 25th and 75th percentile within a peer group you select, such as Shopping then Apparel. The range matters more than the midpoint, which is this post's argument.

My rate is fine but sales are flat. What now?

A healthy blended rate can hide a broken segment, and most sites carry around forty small problems rather than one large one. Work through the conversion leak index to find them, and read what a CRO audit costs before paying anyone.

My store is on Shopify. Where do I look?

Shopify gives you a better instrument than a benchmark. The Conversion rate breakdown report splits sessions, cart additions, reached checkout and completed checkout, so you can see which step leaks rather than judging one blended number. That funnel is walked through in where Shopify stores actually leak.

I run a free trial. Which rate should I even compare?

Not the blended one. A trial has three separate rates, and the 2026 median of 8% describes almost nobody because the distribution has two humps. The trial version of this question splits them.