What is ecommerce conversion rate optimization?

Ecommerce conversion rate optimization (CRO) is the work of getting more of your current visitors to buy. You find where shoppers stall, change that part of the page, and run an A/B test to prove the change earned more sales before you keep it.

It is not a list of tips. An idea that lifts sales on 1 store can cut them on another, and some ideas that look like sure wins lose. The only way to know is to test on your own traffic.

Below, we explain how ecommerce CRO works, what counts as a good conversion rate, and what 4 of our published A/B tests on product, collection and checkout pages showed. 3 won. 1 lost. All 4 taught us something.

Ecommerce CRO gets more buyers from the traffic you already have

Your conversion rate is the share of visits that end in an order. To work it out, divide orders by sessions and multiply by 100. A store with 40,000 sessions and 800 orders in a month converts at 2%.

Now say a test lifts that rate to 2.2%. That is 880 orders from the same 40,000 sessions: 80 more orders a month with no extra ad spend. This is why CRO pays off. Every gain applies to all the traffic you already buy, every month after you implement it.

A detail trips people up. When a test result says “conversion rose 10%,” it means the rate rose 10% of itself, for example from 2.0% to 2.2%. It does not mean the rate jumped from 2% to 12%. All the lifts in this article are measured that way.

A CRO program runs the same loop again and again:

1. Research. Use analytics, heatmaps and session recordings to find where shoppers drop off.
2. Hypothesis. Write down what you will change, what you expect to happen, and which number should move.
3. Test. Split traffic between the current page and the new version, and measure both.
4. Decide. Implement the winner, or pause the test if it lost or did nothing.

If you are comparing conversion rate optimization services, this loop is what you are paying for. A good partner shows you each step, and the losses along with the wins.

A good conversion rate ranges from 0.57% to 5.81% by category

There is no single good conversion rate for ecommerce. It depends on what you sell. IRP Commerce tracks session conversion rates for online retailers in the UK and Ireland. In August 2026, the overall rate was 2.23%. But the spread by category was huge:

• Arts and crafts: 5.81%
• Health and wellbeing: 3.31%
• Kitchen and home appliances: 2.98%
• Fashion, clothing and accessories: 1.86%
• Baby and child: 0.57%

That is a 10-fold gap between the top and bottom categories. A craft supply store at 3% would be well below its peers. A store selling baby gear at 3% would be far above them.

Category is only part of it. Price matters: a $30 refill sells faster than a $900 stroller. So does traffic. Returning customers from email convert far better than first-time visitors from a social ad. A store that doubles its ad spend on cold traffic can see its conversion rate fall, even while total orders rise.

So use benchmarks as a rough check, not a goal. The most useful comparison is your own store: the same pages, the same season, the same traffic sources, before and after a change. That is exactly what an A/B test measures.

Ecommerce CRO works on 4 steps where shoppers drop off

Most online stores have the same path to an order: a landing page, a collection page, a product page, then cart and checkout. Shoppers leave at every step. CRO finds the step that loses the most of them and works there first.

1. Homepage and landing pages

This step has 1 job: get shoppers into the store. Watch for visitors who leave without clicking anything, or who click around the menu and never reach a product. Common fixes to test are clearer category links, less motion in the hero, and a single strong button into the shop.

2. Collection pages

Here shoppers compare and narrow down. Watch for long scrolls with no product clicks, heavy use of filters that return nothing, and quick exits back to the menu. Tests here often cover the grid layout, the filters, what each product card shows, and trust signals near the top.

3. Product pages

This is where most buying decisions happen. Watch for shoppers who view a product and never add it to the cart. Tests often cover how the price is shown, photos, reviews, shipping and delivery details, and the add to cart button.

4. Cart and checkout

Shoppers here have already chosen a product, so everyone who leaves is a sale you almost had. Watch for drop-off at each form step, and for surprise costs like shipping that appear late. Tests often cover shipping messages, form length, payment options and reasons to finish now.

Your analytics funnel report shows how many sessions move from 1 step to the next. The step with the biggest drop is usually the best place to start. The 4 tests below cover 3 of these steps.

A strike-through price lifted product page sales 16.1%

The product page is where most buying decisions get made, so small changes there can move a lot of revenue. On 1 store, the product page showed only the current sale price. Shoppers had nothing to compare it to.

We tested adding the list price with a line through it, next to the sale price, plus the dollar amount saved just below. The idea was simple: make the value clear at a glance.

Conversion rose 16.1% over 29 days, and revenue per visitor rose 16.7%. Average order value held about steady. That tells you where the gain came from: more shoppers buying, not existing buyers spending more.

The takeaway is that shoppers judge a price against something. If you give them no reference point, they find their own, often a competitor’s page. A struck-through price gives them one on your page.

A rule applies here. Only show a list price the product really sold for. A made-up “was” price can break pricing rules where you sell, and shoppers who spot it stop trusting the rest of the page.

Star ratings lifted conversion 11.3% but cut revenue per visitor 7.5%

Collection pages are where shoppers compare and narrow down. On 1 store, collection pages opened with a category heading and a block of intro copy. Trust signals did not show up until much later in the visit.

We tested 2 ways to build trust at the top of the page. Variation A showed a hand-picked customer review with the buyer’s photo. Variation B showed the site’s star rating, matching the rating already in the homepage hero, and moved the intro copy lower.

The single review showed little effect, and we paused it partway through to send more traffic to the star rating. The star rating won on conversion: up 11.3% over 33 days. A quick rating built more trust than 1 review that shoppers had to stop and read.

But revenue per visitor fell 7.5%. More shoppers bought, and each one spent less. The test does not tell us why. It is the open question to watch after implementing, and it is a good example of why 1 number is never enough to judge a test.

When conversion and revenue per visitor disagree, decide which one matters more for the business before you implement. A store that wants more first-time buyers may accept smaller orders. A store with thin margins may not.

A plain checkout timer lifted mobile orders 8.0%

Checkout is the last step, and shoppers who reach it have already decided they want the product. Many still leave. On mobile, they get distracted, switch apps, or plan to come back later and never do.

We tested a countdown timer on mobile checkout that showed how long the cart was reserved. Variation A used a plain timer. Variation B used the same timer with an emoji and a line saying the cart was reserved.

The plain timer won. Conversion rose 8.0% in 18 days, and revenue per visitor rose 15.0%. Average order value rose about 6%. The emoji version trailed. It likely read as less serious at the moment shoppers were about to pay.

2 lessons come out of this test. First, tone matters as much as the idea. The same timer, dressed up differently, did worse. Second, the version you test is the version you learn about. If we had only tested the emoji version, we might have decided timers do not work at checkout.

The winning timer went out to all mobile shoppers through the testing tool while the store built it into the site. Like the price test, the timer showed something real: how long the cart was held. Fake urgency is easy to spot and costs trust.

Every store has more buyers in the traffic it already has. A test shows which change brings them out.

Single product rows on mobile cut conversion 9.4%

Not every good idea wins. On 1 store, mobile collection pages showed products in a grid, 2 across. We tested showing 1 product per row at full width. The thinking was that bigger product cards would help shoppers focus on each product.

It lost. In the single product rows test, conversion fell 9.4% over 14 days, and revenue per visitor fell 7.2%. With 1 product per row, a long collection became too long to browse. Shoppers saw fewer options for the same amount of scrolling. We paused the test and kept the 2-across grid.

A loss like this is still a result. Without the test, this change would have looked like a clear upgrade, and it would have gone live on every collection page. The test caught it in 14 days and cost the store nothing after that.

A testing program that keeps winning follows 5 steps

The 4 tests above came from a program, not a list of ideas. Here is how that program runs.

1. Start where shoppers drop off. When I start with a new store, the first tests come from what we find in the research phase. We look at the conversion rate of each step in the funnel and go after the step with the highest drop-off. If no step stands out, we start with the landing pages that get the most traffic. Our guide on how to conduct a CRO audit walks through that research.
2. Write a clear hypothesis. Every test in our archive starts with the same shape: “By [making this change], we will [cause this effect], resulting in [this number moving].” If you cannot fill in all 3 parts, the idea is not ready to test.
3. Test on the right device. 2 of the 4 tests above ran on mobile only. Shoppers on a phone and on a laptop use the same store differently, so a change that helps 1 can hurt the other. Split your results by device, or test each one on its own.
4. Run full weeks and judge 2 numbers. Weekday and weekend shoppers behave differently, so we run most tests for at least 2 full weeks. The tests above ran from 14 to 33 days. Judge each one on conversion rate and revenue per visitor, never conversion alone.
5. Implement winners, pause losers, and write both down. Each result tells you something about your shoppers. The star rating test, for example, says this store’s shoppers respond to quick trust signals. The next collection page test can build on that.

Run on Shopify? Our guide to Shopify conversion optimization techniques covers tests built for Shopify stores.

Small, clear wins stack up. A 16.1% lift on the product page and an 8.0% lift at checkout work on different shoppers at different steps. Over a year of testing, gains like these add up to far more than 1 big redesign you cannot measure.

Frequently asked questions

What is ecommerce conversion rate optimization?

Ecommerce conversion rate optimization (CRO) is the process of getting more of your current visitors to buy. You research where shoppers drop off, change that part of the site, and A/B test the change to prove it earned more sales before you keep it.

What is a good ecommerce conversion rate?

It depends on your category. IRP Commerce data for UK and Irish online retailers put the overall rate at 2.23% in August 2026, but categories ranged from 0.57% for baby and child to 5.81% for arts and crafts. Your own rate over time is the most useful benchmark.

How do you calculate conversion rate?

Divide orders by sessions and multiply by 100. A store with 800 orders from 40,000 sessions has a 2% conversion rate.

How long should an A/B test run?

Run it for full weeks, long enough to reach a confident result. We run most tests for at least 2 full weeks. The ecommerce tests in this article ran from 14 to 33 days.

What is the difference between conversion rate and revenue per visitor?

Conversion rate counts how many visits end in an order. Revenue per visitor divides total revenue by visitors, so it also reflects order size. They can move in opposite directions: in 1 of our tests, conversion rose 11.3% while revenue per visitor fell 7.5%.