What's a good conversion rate? Benchmarks by industry and traffic source

For an ecommerce store, sitewide, a conversion rate around 2% is normal in 2026 — above 3% is good, above 5% puts you near the top of the distribution. For a lead-generation site, where the conversion is a form fill or a call rather than a purchase, the averages run higher: roughly 3–6% depending on industry. And for a dedicated landing page receiving campaign traffic, the median is 6.6%, with strong pages well into double digits.

Three different answers, and none of them is wrong. Which one applies to you is the actual question, and it matters more than any amount of optimization advice. A store owner who reads a landing-page study and concludes their 2.1% site is failing will start fixing things that were never broken.

Search for 'average conversion rate' and you'll find figures from under 2% to over 8%, most of them quoting each other without saying what was measured. The disagreement comes from two definitions that studies rarely state up front.

The first is the denominator. Sitewide rates divide by every visit to any page — including blog readers, support searches, and people who bounced off the homepage in four seconds. Landing-page rates divide only by visits to one page built for one campaign. Ad-platform rates divide by clicks on an ad. Same business, and the three denominators produce numbers that differ by 3–5x.

The second is what counts as a conversion. A completed purchase is a high bar. A submitted form is a lower one, which is why lead-gen sites post rates an ecommerce store will never see. Some landing-page tools count any click on the call-to-action, which is lower still. There's also a quieter version of this problem — whether 'visitors' means people or visits — that can shift your own rate by a wide margin; the conversion tracking guide covers it.

So before comparing yourself to any number below, check that it counts what you count. That test disqualifies most of what ranks for this query.

Ecommerce is the best-measured category, because the conversion is unambiguous: money changed hands. IRP Commerce publishes a live panel of trading data, and it put the average at 2.03% in June 2026 — up from 1.85% a year earlier. Littledata's Shopify benchmark runs lower at 1.4%, with the top tenth of stores above 4.7%, though that dataset was assembled in 2023 and is still quoted as current everywhere. Dynamic Yield, whose sample skews toward larger brands, reports 2.69%.

Read those together as 1.4–2.7%, with small stores at the low end and established brands at the high end.

What you sell moves the number more than how well you sell it. In IRP's June 2026 sector data, arts and crafts converted at 5.53% while fashion sat at 1.70% and baby products at 0.51% — a tenfold spread inside the same month, same methodology. Cheap, frequently repurchased goods convert well; expensive, comparison-heavy purchases convert badly, because visitors do research across several sites before buying on one of them. If your rate looks poor against 'the average', check the sector figure before concluding anything. And wherever your rate stands, the biggest known leak sits at the end of the funnel: about 70% of carts are abandoned before checkout completes.

When the conversion is an enquiry rather than a sale, the bar is lower and the rates are higher. The most useful current dataset is Ruler Analytics' 2026 benchmark — 110 million sessions and five million conversions across 13 industries, where a conversion means a qualified lead or sale. Overall average: 5.13%.

The industry spread is what makes the average nearly useless on its own. Automotive and legal lead at 7.9%, software at 7.6%, with finance and education around 6.3%. At the other end: real estate at 2.8%, health care at 2.3%, travel at 1.9%. Ruler's retail and ecommerce figure, 2.4%, lands close to IRP's independently measured 2.03% — two unrelated datasets agreeing is about the strongest confirmation this field offers.

One caution for B2B SaaS specifically: First Page Sage's benchmarks put visitor-to-lead conversion at just 1.1% for SaaS sites against 7.4% for legal services. A SaaS homepage serves researchers, comparison shoppers, and existing users logging in; a personal-injury firm's site serves people who need a lawyer this week. Intent explains most of the gap, and no amount of button-color testing closes it. (If you're measuring a SaaS funnel past the website — trial-to-paid and the like — that's a different metric with its own benchmarks.)

The double-digit figures that make sitewide rates look sickly come from studies with a narrower denominator. Unbounce's Conversion Benchmark Report, built from 41,000 dedicated landing pages and 464 million visits, puts the median at 6.6% — ranging from 3.8% for SaaS to 12.3% for events and entertainment. These pages receive campaign traffic, make one offer, and count a form fill as success. Of course they convert better than a whole website; that's the job.

Ad platforms measure from a click, which is narrower again. WordStream's 2026 Google Ads benchmarks, from 13,474 US campaigns, average 8.18% — someone who clicked a search ad had already typed a commercial query, the highest-intent traffic that exists. The spread runs from 16.22% for pet services down to 2.64% for finance and insurance.

These numbers are for judging campaigns and landing pages, not websites.

The same site converts its visitors at very different rates depending on where they came from. In Ruler's 2026 channel data, paid search converts at 5.4%, with email and organic search at 4.9%, referral at 4.8%, and direct at 4.7% — a tight cluster. Then comes social: 2.2% organic, 2.1% paid. Roughly half the rate of everything else, and the pattern holds across studies and years. A person mid-search has intent; a person mid-scroll was interrupted.

The 2026 curiosity at the top of that table: visitors arriving from AI assistants — ChatGPT and its relatives — convert at 5.8%, the highest of any channel. The volume is still small, but someone who asked an AI for a recommendation and clicked through arrives with the question already answered.

The practical consequence: your traffic mix can move your sitewide rate while nothing on your site changes. A post that goes viral on social will drop your conversion rate that week, and it's good news. Comparing your rate month to month without segmenting by source is comparing two different audiences — which is why tagging your campaign links comes before any serious conversion analysis.

A benchmark answers exactly one question — is my number in the plausible range for my industry and traffic mix? — and it answers it once. If you're an online store at 1.9% or a software site at 6%, you're normal, and the benchmark has told you everything it knows.

Past that check, the comparison that pays is against your own site: this month against last month, this source against that source, this landing page against the other one. Those numbers share your definition, your counting quirks, and your audience, so a change in them means something real changed. A gap against an industry average could mean a hundred things, most of which you can't act on.

Two habits keep the self-comparison honest. Segment before reacting — a falling sitewide rate with every channel steady means your mix shifted, and that diagnosis takes one look at the source report. And respect small numbers: at eleven sales a month, a swing to nine is noise, and the fix for noise is tracking the smaller conversions upstream that happen often enough to show a trend.