What Is a Good Funnel Conversion Rate? Sourced Benchmarks and How to Find Your Own

The short answer: As of October 8, 2026, there is no single good funnel conversion rate, and most numbers quoted online do not say what they measured. The few benchmarks with a named sample and definition: Contentsquare's 2026 data (99 billion sessions, 6,000+ sites) shows 2.9% of returning visits and 1.7% of new visits include a conversion event; Unbounce's 2024 report puts the median landing page at 6.6%; Baymard's average cart abandonment is 70.22%; ChartMogul's January 2026 survey puts median SaaS free-to-paid at 8%. A good rate for you is one that beats your own baseline for the same steps and the same traffic source.

TL;DR

• "Funnel conversion rate" means different things depending on the first step, the last step and the denominator (visits or people). Benchmarks only help when those match yours.
• Few published benchmarks state their sample, date and conversion action. The four in the table below do. We left out the rest, including the widely repeated B2B stage ranges.
• Compute step rates, not just the overall rate. The overall rate tells you how much leaks; step rates tell you where.
• Segment by traffic source before comparing periods. A mix shift toward lower-intent traffic lowers the blended rate even when nothing on the site changed.
• Set your baseline from at least four weeks of your own data, then judge changes against it.

What "funnel conversion rate" actually measures

A funnel is an ordered set of steps, such as landing page view, product page view, add to cart, checkout started and purchase. Two rates come out of it:

Three choices change the number before any visitor does anything:

1. Where the funnel starts. Starting at the landing page includes every bounce. Starting at the product page excludes them. The same store can report 2.1% or 4.7% depending on this one choice (see the worked example below).
2. What the last step is. A form submission, a free trial signup and a paid order are all "conversions", and they sit at very different rates.
3. The denominator. Visits (sessions) and people (users) give different answers. One person who visits three times and buys once is a 33% visit conversion and a 100% user conversion. Dynamic Yield's formula divides actions by users; Contentsquare's divides by visits. Neither is wrong, but you cannot compare one with the other.

If a benchmark does not tell you all three, you cannot tell whether your number is good or bad against it.

The benchmarks that pass a sourcing check

We only include a figure here if the publisher names the sample, the date and the conversion action. Everything was checked on the publisher's own page on October 8, 2026.

Read these with their definitions attached:

• Contentsquare's rates are per visit and mix action types. A purchase and a newsletter form both count. Use them for directional comparisons by source and visitor type, not as a target for a checkout funnel. Contentsquare also reports desktop converting 74% higher than mobile, so a mobile-heavy site should expect a lower blended rate.
• Unbounce's 6.6% is a single landing page step, mostly lead capture and click-throughs on pages built for campaigns. It is not a purchase rate and not a full funnel rate. Comparing your visitor-to-order rate with it will make any store look broken.
• Baymard's 70.22% is an average of other people's studies, so each underlying study has its own method. It is the best-known reference for the cart-to-order part of an ecommerce funnel. Baymard's separate 2026 survey of 1,083 US online shoppers lists "extra costs too high" as the top fixable reason for abandoning, cited by 40%.
• ChartMogul's figures are survey answers about a six-month window, so they run higher than a 14-day trial-to-paid rate measured inside your billing tool.

What we left out, and why

Several numbers on the pages that rank for this query did not pass:

• B2B stage rates such as lead to MQL, MQL to SQL and opportunity to close. The most cited source describes its data as internal and anonymized client data from 2017 through 2025, without stating how many companies or leads it covers. Several other pages repeat these ranges second-hand.
• "Most funnels convert between 3% and 10%" and "10%+ is excellent." We could not trace either to a primary dataset.
• Global ecommerce purchase rate dashboards that do not state the sample size. One widely linked dashboard was quoted at 3.17% on a top-ranking page; the same dashboard showed 2.68% on October 8, 2026, because it is a rolling twelve-month figure. Rolling figures move, which is one more reason to quote a source with its date.
• Single-practitioner quotes ("a good visit-to-lead rate is 3.1% to 5%"). Useful color, not a benchmark.

This is the same rule Humblytics applies on its calculators: a benchmark without a stated sample and conversion action does not get published.

Why "good" depends on traffic source, price and step definition

Even a well-sourced benchmark only fits if your funnel looks like the sample. Four variables move the number more than page design does:

Worked example: one store, three "conversion rates"

The numbers below are hypothetical, chosen to show the method. They were computed with a short script; percentages are rounded to one decimal place.

A store gets 20,000 landing page visits in a month from two sources: 14,000 from paid social and 6,000 from email.

What the table shows:

1. The same store has three honest answers. Landing page to purchase is 2.1%. Product page to purchase is 4.7% (420 ÷ 9,000). Checkout started to purchase is 60.0%. Each is "the funnel conversion rate" depending on who is asking.
2. The biggest leak is product page to add to cart. At 15.6% blended, it loses 7,600 people, far more than any later step. That is where a test belongs, not the checkout.
3. Compare like with like. The cart-to-order step here loses 70% of carts (1 minus 420 ÷ 1,400). That is the number to set beside Baymard's cart abandonment average, because both measure carts that do not become orders. Setting the 2.1% overall rate beside Unbounce's 6.6% landing page median would be meaningless.
4. A mix shift can fake a decline. Next month, paid social doubles to 28,000 visits and every step rate in both columns stays exactly the same. Orders rise from 420 to 600, but the blended overall rate falls from 2.1% to 1.8%. Nothing on the site got worse. Report this funnel blended and you would start "fixing" pages that work.

To run the same arithmetic on your own counts, use the conversion rate calculator for each step.

How to find your own baseline

A baseline is your own funnel's rate over a stable period, with the definitions frozen. It is the only benchmark guaranteed to match your steps, price and audience.

1. Write down the steps. Name the first and last step, whether each step is a page view, a click or a form submit, and whether people must hit the steps in order.
2. Pick the denominator. Visits or people. Keep it fixed.
3. Pull at least four full weeks, so weekday and weekend patterns both appear, and avoid windows with a sale or a launch unless that is the funnel you want to measure.
4. Split by source and by new versus returning. Record a baseline per segment, not one blended figure.
5. Note the noise. Small counts swing. In the example, 420 purchases from 20,000 visits gives a 95% range of about 1.9% to 2.3%. A single week of 105 purchases from 5,000 visits gives about 1.7% to 2.5%. A one-week dip inside that range is not a finding.
6. Re-baseline when the definition changes, not when the number disappoints.

How to compare like with like

Before you compare your rate with anything, check five things:

If more than one row does not match, compare against your own history instead. If you want a structured list of what to fix once you know the leaking step, the ecommerce CRO checklist covers it step by step, and the A/B test hypothesis guide shows how to turn a drop-off into a testable change.

Measuring step-by-step funnels in Humblytics

For the background on why step analysis matters, see what funnel analysis is and how it drives growth. For the setup itself, see the complete guide to funnel analysis in Humblytics.

In short, Humblytics funnel steps can be a page view, a click or a form submit, run in sequential or any order, and can span domains, so a funnel that starts on a marketing site and ends on a checkout subdomain counts as one funnel. Each funnel reports in-funnel and off-path visitors, percent continued, drop-off and entry point, with flow and Sankey views. When the last step is a payment, revenue comes from Stripe or Foxy, so a funnel can be judged on dollars rather than order counts. The Business plan includes 5 funnels; Scale includes unlimited funnels.

When you find the leaking step, you can test a fix on the same script. Significance is calculated in real time, tests can stop after a set duration, and promoting a winner stays a human decision. See pricing or start a 14-day trial.

Sources and freshness

• Contentsquare, Conversion benchmarks: tracking growth in a high-cost, high-intent world, retrieved October 8, 2026. Supports the 99 billion sessions and 6,000+ sites sample, returning (2.9%) and new (1.7%) visitor rates, paid search (2.8%), AI-referred (1.3%, +55%, Q4 2024 to Q4 2025), organic social (0.7%) and desktop 74% higher than mobile.
• Contentsquare, Digital Experience Benchmark 2026, retrieved October 8, 2026. Supports the conversion rate definition (share of visits that include a conversion event, like a purchase or form submission).
• Unbounce, Conversion Benchmark Report 2024, retrieved October 8, 2026. Supports the 6.6% median, the 3.8% to 12.3% industry range, and the sample of 41,000+ landing pages, 464 million+ visitors and 57 million+ conversions.
• Baymard Institute, Cart Abandonment Rate Statistics, retrieved October 8, 2026. Supports the 70.22% average across 50 studies, the September 22, 2025 update date, and the 2026 survey of 1,083 US adults (extra costs too high, 40%).
• ChartMogul, The SaaS Conversion Report, retrieved October 8, 2026. Supports the January 2026 survey of 200 B2B products, the 8% median free-to-paid rate, the 30% credit-card trial rate and the six-month conversion window.
• Dynamic Yield, eCommerce conversion rate benchmarks, retrieved October 8, 2026. Cited only for its users-based formula and its rolling twelve-month figure (2.68% on the retrieval date); not used as a benchmark because the page does not state its sample size.

Benchmark reports are updated yearly or roll monthly. Check the linked pages before quoting a figure, and quote it with its date.