- A good conversion rate is relative. The most useful comparison is your own rate from the same period last year, measured in the same way.
- Benchmarks depend on how they were measured: per session or per user, which shops, which year. A value without a source and a survey year is not a benchmark.
- Cookie consent shrinks the measurement base. A rate calculated only on visitors who accept cookies is not directly comparable with a benchmark that uses a different measurement base.
- If your rate is low, find the stage where buyers drop off before you change the design, then test the fix and measure it in your own GA4.
A good conversion rate for an online shop depends on the industry, the price level, the device and the traffic source. There is no single number that fits every shop. The most reliable comparison is your own conversion rate from the same period last year, measured with the same definition and split by device and source.
This page explains how to read benchmarks without being misled. It shows benchmark values by industry and by device from one documented source: Littledata's 2026 analysis of 433 Shopify stores that measure with Google Analytics 4. It also shows how to calculate your own rate correctly, why measurement differences distort comparisons, and where to start when your rate is below what you expected.
Short answer
There is no single good conversion rate for online shops. The rate depends on the industry, the price level, the device, the traffic source and the way it is measured, so one number from another shop rarely fits yours. The most useful comparison is your own rate from the same weeks last year, measured the same way and split by device and traffic source.
Published benchmarks are useful as a rough orientation, if you know how they were collected. The figures on this page come from one source, the Littledata ecommerce conversion rate benchmark: 433 Shopify stores that send their data to Google Analytics 4, over the 90 days from July 20 to October 17, 2026, measured per session. In that sample, the median store converts 1.4% of sessions into orders, and the top 20% of stores convert more than 2.7%.
Use the sections below in this order: look up the benchmark for your industry, check it against your device mix, then read why such numbers mislead before you draw a conclusion.
Benchmarks by industry
The table shows the median conversion rate per session by industry, and the rate a store needs to be in the top 20% of its industry.
| Industry | Median conversion rate | Top 20% of stores above | Stores in sample |
|---|---|---|---|
| All Shopify stores | 1.4% | 2.7% | 433 |
| Beauty & Skincare | 2.9% | 4.0% | 38 |
| Health & Supplements | 2.4% | 3.8% | 35 |
| Food & Beverage | 2.4% | 3.2% | 31 |
| Fashion & Apparel | 1.1% | 2.0% | 71 |
| Home & Furniture | 1.0% | 1.7% | 62 |
Source: Littledata, ecommerce conversion rate benchmarks. Shopify stores using Google Analytics 4 with at least 1,000 sessions and 90 purchases, data from July 20 to October 17, 2026, retrieved October 21, 2026. Industries are listed only where the sample is large enough.
When you read a benchmark, check three things: how the rate was defined (per session or per user), which shops were included (size, region, price level), and which period it covers. A value from a different definition is not comparable, even when the label looks the same.
Benchmarks by device
Conversion rates differ between phones, tablets and desktop computers, and between traffic sources such as search, paid ads, email and direct visits. Part of the gap comes from intent: a visitor who types a product name into a search engine is closer to buying than a visitor who clicks a social post. Part of it comes from the device: small screens make forms harder, and many people research on a phone and buy later on another device. The same source reports conversion rates for phones and desktop computers. It gives no separate tablet value and no figures by traffic source.
| Device | Median conversion rate | Top 20% of stores above | Top 10% of stores above |
|---|---|---|---|
| Mobile | 1.6% | 2.9% | 4.1% |
| Desktop | 1.5% | 3.1% | 5.1% |
Source: Littledata, ecommerce conversion rate benchmarks, data from July 20 to October 17, 2026. Device figures include only stores with at least 500 sessions and 30 orders on that device. In this sample the medians are almost equal; the gap opens among the strongest stores, where the top 10% reach 5.1% on desktop and 4.1% on mobile.
You can build the same table for your own shop in minutes, and add traffic sources to it. Open your analytics tool, split the conversion rate by device category and by source, and write the numbers down for each month. That table is the one that matters for decisions.
Why benchmarks mislead
A benchmark is an average over shops that differ from yours. Four effects make it mislead:
- Different definitions. A session is one visit. A user is one person who may visit several times. The same shop has a lower rate per session than per user, because repeat visits that end without an order count as extra sessions. Compare like with like.
- Consent shrinks the measurement base. If your analytics only records visitors who accept cookies, your rate is calculated on a subset of visitors. That subset usually behaves differently from the full audience, so your rate is not directly comparable with a benchmark that uses a different measurement base.
- Assortment. A shop with a few well-known products and a shop with a large range of rarely bought items sit at different levels for good reasons.
- Price level. Higher prices usually mean longer decisions and more visits before an order. A fair comparison needs shops with a similar basket.
A fifth effect comes from outside: a campaign, a sale or a news event can lift or lower the rate for weeks. A comparison with last year only works for the same season and a similar marketing mix. For that reason, treat a benchmark as a question to ask of your data.
A responsible way to use one: write down where the number comes from, check that your measurement matches its definition, compare it with your rate for the same segment (never your overall rate), and note the gap as a hypothesis to investigate. A gap of any size is a reason to look at your data, and it is no proof of a problem.
How to calculate your own rate correctly
The conversion rate is the number of orders divided by the number of sessions (or users) in the same period, multiplied by 100:
Conversion rate (%) = orders ÷ sessions × 100
The blog guide to the conversion rate explains the formula in detail and includes a calculator. For a number you can trust, follow these rules:
- Pick one denominator and keep it. Sessions are the common default in analytics tools. If you switch to users, switch for all comparisons.
- Count orders once. Use the purchase event, and check that reloads of the confirmation page do not create duplicates.
- Remove noise. Exclude internal traffic and known bots, and keep the same filters over time.
- Use the same period and season. Compare full weeks, and compare with the same weeks a year earlier.
- Split before you conclude. Report device and source separately, and keep the overall rate as a summary.
Check also that your analytics data and your shop system agree on the number of orders. If they differ by a lot, fix the tracking before you compare anything with a benchmark.
Below benchmark? Where to start
If your rate sits below the figures you find, do not start with design changes. Start with the cause. The diagnosis in Why your online shop is not converting follows the buying path stage by stage and shows where visitors drop off. The overview of conversion optimization covers the method and typical measures.
To check that a change really improved your rate, run it as an A/B test. Here is how Varify.io supports that:
- Test ideas for a page. The AI CRO Audit takes the URL of a page and returns CRO insights, recommendations and ten test ideas you can use directly as A/B tests.
- Describe the change, let the AI build it. With Prompt-Based Experimentation (PBX) you describe the experiment in plain language, and the AI builds the variation and sets up the experiment, without a developer. See PBX examples.
- Measurement in your own GA4. Experiment data and analytics data come from the same source, so the numbers in your test report and in your analytics tool match. See tracking with Varify.io.
- No cookies. Varify stores the variation assignment in localStorage, without personal data. More on cookieless A/B testing.
- A readable result. The free significance calculator shows whether a measured difference is likely to be real.
Start with one change, a clear goal and a fixed runtime. When the test ends, compare the result with your own rate from before, not with a benchmark from someone else.
Check your conversion rate against your own history.
Measure experiments in your own GA4 and see which change moves your rate.