Careers
Part of Conversion rate optimization: parts worth your attention
Conversion rate optimization benchmarks: the numbers and what sits behind them
Conversion rate optimization benchmarks in 2027 compare one defined state transition, stable counts, customer quality, task access, and feasible test targets.
What to take away
- Sitewide ecommerce conversion in the United States typically lands between 2 percent and 3 percent, and most product categories sit inside a 1 percent to 4 percent band.
- Paid search commonly converts 2 percent to 5 percent of clicks, while display and cold social traffic usually run between 0.3 percent and 1.5 percent.
- Cart abandonment has averaged near 70 percent for years, so checkout holds more recoverable revenue than the homepage does.
- Published averages are a check on your own data, not a target to copy.
The title asks for numbers, so start with the ranges. A healthy sitewide ecommerce conversion rate in the United States usually falls between 2 percent and 3 percent. Above 4 percent is strong for general merchandise. Below 1 percent usually means tracking needs checking before design does.
These are typical ranges from commercial benchmark publishers, not audited figures. Each publisher counts a conversion differently, and the counting method matters more than the decimal. The definitions behind those counts are the conversion rate optimization foundations worth fixing before you compare yourself with anyone.
Where the published numbers land
| Channel or funnel step | Typical conversion range | What it counts | Commonly cited source |
|---|---|---|---|
| Ecommerce sitewide, US | 2% to 3% | Orders divided by sessions | Contentsquare and Dynamic Yield quarterly reports |
| Paid search landing page | 2% to 5% | Orders or form completions divided by clicks | Google Ads benchmark data, aggregated by agencies |
| Display and cold social | 0.3% to 1.5% | Same actions on prospecting traffic | Platform and agency benchmark roundups |
| Email campaign | 1% to 3% | Orders or leads divided by delivered messages | Mailchimp and Klaviyo benchmark reports |
| B2B visitor to lead | 1% to 3% | Identified leads divided by sessions | HubSpot benchmark data |
| B2B lead to qualified opportunity | 15% to 30% | Accepted opportunities divided by leads | Sales and marketing benchmark studies |
| Checkout, cart to completed order | 25% to 35% | Orders divided by carts created | Baymard Institute checkout research |
Every row changes if the denominator changes. Orders per session, orders per user and orders per cart give three different rates from the same week of data.
Email is where deliverability distorts the benchmark most. If a tenth of the list never reaches the inbox, conversion per delivered message looks healthier than revenue does, and inbox placement moves the number more than subject lines do. Email deliverability sets out which metrics publishers report.
Period matters as much as source. Most of these publishers refresh quarterly or annually, so a rate quoted without a date is a rate you cannot use.
What sits behind the averages
Baymard Institute compiles published cart abandonment studies, and its average has sat near 70 percent for years. Its own usability testing places much of that loss at checkout, on forced account creation and on costs that appear late. That is a process number, not a persuasion number.
Commercial ecommerce reports have clustered near 2 percent to 3 percent sitewide for several years. The stability is the useful part. Any claim that a redesign doubled conversion should be checked against the same denominator, the same traffic mix and a mature outcome window, which is where many conversion rate optimization mistakes begin.
Splits that change the decision
Split before you judge. Desktop ecommerce conversion commonly runs one to two percentage points above mobile in the same store, and returning visitors often convert several times better than first-time visitors. Check these cuts on your own data first.
- Devicemobile against desktop for the same period.
- Traffic sourcepaid, organic, email and direct.
- Visitor typenew against returning.
- Funnel stepcart to checkout, checkout to order.
- MarketUnited States against Canada, where you sell in both.
Averages also hide access problems. A rate built only from analytics counts the people who abandoned at the first field, not the people who could not complete it. Task completion and error counts from repeated testing rounds, run with the GOV.UK usability benchmarking method, give you a second benchmark that analytics cannot produce.
Example: one store, three numbers
Same store, same quarter: 3.2 percent sitewide, 1.9 percent on mobile, 2.7 percent once promotion traffic is excluded.
Only one of those three numbers describes the design. The sitewide rate is diluted by paid traffic from a promotion. The mobile rate is the one a design team can act on this month. Publishing all three ends an argument before it starts.
Decide in advance which number you will report, then hold that choice for the whole series. The conversion rate optimization checklist covers the definitions that need locking first.
Build the range you steer by
- Pick one state transitioncart to completed order, or visit to accepted lead.
- Fix the denominator and write it downsessions, users or carts.
- Take at least four comparable periods that cover a full decision cycle.
- Record counts and revenue beside every rate, not percentages alone.
- Note what explains outliersprice changes, outages, promotions, stock.
Set targets from value, not from an average. Work out the smallest effect worth the build cost, then size the test from your own baseline. A move in the rate after a change is not automatically caused by that change, and the GAO evaluation design guide is a reminder that evidence needs a design behind it.
Once the range is stable, improving conversion rate optimization starts with the worst step in the funnel you defined, not with the homepage.
