Ecommerce conversion rate: what's good, how to measure it and how to raise it
The conversion rate formula, public benchmarks by category, the most common measurement mistakes and the levers to pull at each funnel stage.

Ecommerce conversion rate is the percentage of sessions (or visitors) that end in a paid order. The math is orders divided by sessions, times 100. Public benchmarks tend to land around 2%, but they swing widely by category, traffic source and device. The most useful comparison is always your own store's history.
This guide is part of our topic on ecommerce conversion rate optimization and A/B testing. It covers the formula, what should count as a conversion, which market numbers exist and how to use them without fooling yourself, the most common measurement mistakes and the levers that raise conversion at each funnel stage.
How do you calculate ecommerce conversion rate?
The formula is simple:
Conversion rate = (paid orders ÷ sessions) × 100
Example: 10,000 sessions and 180 paid orders in a month give you a 1.8% conversion rate. The choices that change the result sit on both sides of the division:
- Sessions or visitors? One person may visit three times before buying. Per session, the rate comes out lower; per unique visitor, higher. Both are valid as long as you always use the same one.
- Orders placed or orders paid? Some payment methods create the order before the money arrives: bank transfers, Pix in Brazil, cash on delivery. A placed order that never gets paid is not a sale. For business decisions, count paid orders.
- Which period? Use full weeks at a minimum. A promo day or a holiday skews the average.
What is a good ecommerce conversion rate?
There is no single number, and be wary of anyone who quotes one without saying where it came from. One of the few public benchmarks updated monthly with an explicit definition is the IRP Commerce market data dashboard, which aggregates UK retailers and defines conversion as transactions divided by sessions. In September 2026 the average was 1.97% (checked on November 3, 2026). The spread across categories in the same month shows why an average alone misleads:
| Category (IRP Commerce, Sep 2026) | Session conversion rate |
|---|---|
| Arts and crafts | 5.76% |
| Health and wellbeing | 2.83% |
| Kitchen and home appliances | 2.59% |
| Pet care | 2.13% |
| Fashion, clothing and accessories | 1.62% |
| Food and drink | 1.28% |
| Baby and child | 0.44% |
Three caveats when you use these numbers:
- It is one market. These are UK retailers, with their own buying habits and payment mix. If you sell in the US or elsewhere, treat it as an order of magnitude, not a target.
- Traffic source changes everything. A store living on retargeting and email converts far better than one buying cold prospecting traffic. Compare campaign to campaign, not your store to a market average.
- Price point matters. A $30 product and a $1,500 product have very different decision cycles. A low rate on a high AOV can earn more than a high rate on a low one.
The useful question is not "is my rate good?" but "is my rate going up or down, by campaign and by device?".
Why is your store's rate different from the average?
- Cold or warm traffic: people who have never heard of your brand convert less than past visitors or customers.
- Device: browsing and checkout are harder on mobile, which usually converts lower than desktop even while bringing more visits.
- Landing page: ads that go straight to the product tend to convert better than ads that land on the homepage.
- Offer and price: shipping cost, delivery time and payment terms change the decision at the last step.
- Measurement: a duplicated pixel, bot-inflated sessions or unpaid orders counted as sales push the number up or down.
Which measurement mistakes give you the wrong rate?
Before trying to raise conversion, make sure you are measuring it correctly:
- Counting placed orders as sales. Separate paid orders from pending, canceled and refunded ones.
- Comparing different tools. Your ad manager, your analytics tool and your store platform count sessions and conversions differently. Pick one source of truth for your conversion rate.
- Lost or duplicated purchase events. Ad blockers and browser privacy features drop part of your pixel events. Our guide to server-side tracking explains how to send events from the server and deduplicate them.
- Mixing in promo periods. A week with a big sale is not the baseline for comparing the following month.
How do you measure conversion at each funnel stage?
Overall conversion is the end result. To know where to act, also measure the step from one stage to the next:
| Stage | Rate | If it is low, look at |
|---|---|---|
| Visit → product | Sessions that view a product ÷ sessions | Ad-to-page match, page speed |
| Product → cart | Add-to-carts ÷ product views | Offer, photos, social proof, price with context |
| Cart → checkout | Checkouts started ÷ add-to-carts | Shipping and delivery revealed late, confusing cart |
| Checkout → purchase | Paid orders ÷ checkouts started | Form length, payment options, trust, declines |
In Horsee, UTM Analytics shows this funnel (visits, cart, checkout and purchase) and the conversion rate by source, medium and campaign, along with real revenue and paid orders for each segment. The Horsee Help Center shows how to drill down through the table.
How do you raise your store's conversion rate?
The order that usually delivers the most for the least effort:
- Fix what is broken. A slow mobile page, a hidden buy button, a payment method that fails. None of that needs a test.
- Match the ad to the page. The creative's promise has to appear at the top of the page people land on.
- Answer doubts before they turn into abandonment. Delivery time, shipping, returns and guarantee near the buy button.
- Cut checkout friction. Fewer fields, guest checkout, cards and wallets on the same step. Our guide to on-site checkout vs hosted checkout goes deeper on this.
- Recover shoppers who leave. Some of them come back with a reminder at the right moment. See the sequence in abandoned cart recovery.
- Test the offer. Bundle versus single unit, free shipping over a threshold, an extended guarantee. This is where an A/B test helps you decide with paid-order data.
When should you use A/B testing to raise conversion?
When there is real doubt between two versions and enough traffic to collect sales on both within a few weeks. In Horsee, an A/B test starts from a store page: variant B is served at the same URL, traffic is split at the edge and each version's conversion is counted on paid orders. By default, the verdict waits for at least 7 days and 30 conversions per version before naming a winner.
How should you track conversion rate over time?
A single number says little. What helps you decide is the trend:
- Week over week, always with complete weeks, so you see direction without the noise of one bad day.
- By campaign, because a new prospecting campaign can drag down the store average with nothing wrong on the page.
- By device, because a mobile-only problem disappears inside the overall average.
- With annotations for everything that changed: a new page, new shipping rates, a new campaign, a payment provider switch. When the rate moves, you know where to look.
A sudden overnight drop is almost always technical: a checkout error, a payment method down or a purchase event that stopped arriving. A slow slide over several weeks is usually market-driven: tired creatives, a competitor with a better price or a shift in your campaigns' audience.
Conversion rate alone is not enough
A store can raise conversion by discounting and still lose money. Track conversion alongside two other numbers: average order value (revenue ÷ orders) and revenue per visitor (revenue ÷ visitors). If conversion climbs while revenue per visitor falls, the change traded margin for volume. For anyone buying traffic, the number that settles the account is cost per paid order compared with the margin on each order.
Measure and test where the sale happens
In Horsee, the funnel by campaign, page-level A/B testing, checkout and cart recovery are on the same platform, and conversion is measured on paid orders. See how it works on the Horsee growth page: native A/B testing and attributed revenue.
Frequently asked questions
How do you calculate an online store's conversion rate?
Divide paid orders by sessions for the same period and multiply by 100. With 10,000 sessions and 180 paid orders, the rate is 1.8%.
What is the average ecommerce conversion rate?
It depends on the source and the category. The IRP Commerce dashboard, which covers UK retailers, reported 1.97% per session for September 2026, with categories ranging from under 0.5% to over 5%.
Is a 1% conversion rate bad?
Not necessarily. With cold prospecting traffic, a high price point or mostly mobile visits, 1% can be normal. Compare against your own history and by campaign.
Should mobile and desktop conversion rates be tracked separately?
Yes. Mobile usually brings more paid-traffic visits and converts lower, so a mobile-only problem can hide inside the overall average. Look at both every week.
What raises conversion rate fastest?
Fixing technical problems on mobile and matching the landing page to the ad's promise. After that come a lower-friction checkout, cart recovery and offer tests.
