By Mehdi HassanWhat are the most reliable landing page statistics?
The most reliable landing page statistics come from a named primary source that states what it measured, on how much data. Many roundups list dozens of figures and copy them from each other. This page lists fewer, because we only include numbers we could trace to an original source, and we flag the popular ones we couldn't.
Each figure below says who published it and what it measures. Where a number is often misquoted, we give the original version. For how to use these figures when you judge your own page, read our guide to landing page conversion rate benchmarks.
- Median landing page conversion rate, all industries (Unbounce)
- 6.6%
- Approximate median for SaaS landing pages (Unbounce)
- 3.8%
- "Watch a Demo" vs "Get Started" CTA (Optimizely test)
- 1.59% vs 0.91%
What do conversion rate statistics actually say?
Conversion rate statistics say less than they seem to, because each one measures a specific stage, page type and traffic mix. The best-documented figures come from Unbounce, which publishes medians from the landing pages on its platform. Visitor-to-lead figures from other sources vary widely.
Statistic | Source | What it measures | Caveat |
|---|---|---|---|
6.6% median conversion rate | Unbounce conversion benchmark report | Landing pages across all industries | Pages built on one platform, mostly single-goal |
About 3.8% median for SaaS | Unbounce conversion benchmark report | SaaS landing pages | Not a sitewide or trial rate |
1.5% to 5% visitor-to-lead | Range across B2B SaaS benchmark sources | Leads divided by visitors | Sources define leads and visitors differently |
These numbers measure different things, so they shouldn't be averaged or compared directly. Our breakdown of SaaS conversion rate benchmarks shows which stage each one belongs to and which widely shared figures lack a method.
Which page speed statistics hold up?
The page speed statistics that hold up are smaller and more specific than the versions people repeat. The widely quoted claim that "one second of delay costs 7% of conversions" traces back to SOASTA data, but the original figures were about 7% for a 100-millisecond delay and about 20% for a full second. A controlled SiteSpect test found a smaller effect: a 4.4% drop on mobile.
Statistic | Source | Caveat |
|---|---|---|
About 7% fewer conversions from a 100ms delay | Original SOASTA data, as reported by MediaPost | One dataset; the effect on your site may differ |
About 20% fewer conversions from a 1s delay | Same SOASTA data | Often misquoted as 7% per second |
4.4% conversion drop on mobile | SiteSpect controlled test | One controlled test, specific setup |
The honest summary: speed affects conversion, the size of the effect depends on the site, and you should measure your own. Our guide to whether website speed affects conversions covers what the evidence supports.
What do CTA and form statistics show?
CTA and form statistics show that specific labels tend to beat vague ones, but the hard data is limited to a few tests. The best-documented example is an Optimizely test where a "Watch a Demo" button converted at 1.59%, against 0.91% for "Get Started". That's one test on one site, so it shows a direction, not a lift you should expect.
Practitioner sources also agree that "Request a demo" and "Submit" are among the weakest CTA labels, though that's an expert consensus rather than a measured figure. Our guide to demo request conversion rate shows how labels fit into the whole booking flow.
Form statistics are where myths cluster. The often-shared claims that multi-step forms convert 86% or 300% better trace back to single case studies or unsourced posts, not to controlled research. Multi-step forms can help when they make a long form feel shorter, but no reliable figure tells you by how much.
What do A/B testing statistics mean for small sites?
A/B testing statistics mean most startup landing pages can't run valid tests. Detecting a 15% lift at a 2% baseline conversion rate takes roughly 15,000 to 20,000 visitors per variant. Many early-stage pages don't get that in a quarter, so a test either runs for months or ends with a result that's mostly noise.
For low-traffic pages, research-led changes work better: session recordings, short on-page surveys, customer interviews and a careful read of the copy. Our page on landing page optimization services describes what that looks like week by week.
What do landing page cost statistics show?
Landing page cost statistics show very wide ranges, because the pages ranking for cost searches quote everything from a template to a full agency build. Typical ranges we see quoted are $500 to $25,000 for a landing page, $1,000 to $50,000 for a CRO audit and $2,000 to $35,000 a month for CRO retainers. These vary with scope and who does the work.
Item | Typical range quoted | What drives the spread |
|---|---|---|
Landing page | $500 to $25,000 | Template vs custom, copy included or not, build platform |
CRO audit | $1,000 to $50,000 | Automated checklist vs research with interviews |
CRO retainer | $2,000 to $35,000 a month | Testing volume, team size, development included |
For reference, our own landing page package is a fixed $6,000, covering copy, custom design and a Next.js build.
Which landing page stats are myths?
The landing page stats most likely to be myths are round, dramatic numbers with no study attached. If a figure appears on dozens of blogs but none links to a primary source, assume it's recycled.
- "Every second of delay costs 7% of conversions." A misreading of SOASTA data, where 7% referred to 100ms.
- "Multi-step forms convert 86% (or 300%) better." Traces to single case studies or unsourced claims.
- "Customer logos lift conversion 43%." We found no traceable primary source.
- "The average landing page converts at X%." Without a stage, page type and source, a single average means little.
How to check a statistic before you cite it
Search for the exact number with the word "study" or "report" and follow the links back until you reach a page that ran the research. Note the year, the sample, the page type and what counted as a conversion.
If you can't reach a primary source within three or four hops, don't use the number. If the original says something different from the version you found, cite the original. Our landing page audit checklist uses the same habit: question the claim before changing the page.
How should you use landing page statistics?
Use landing page statistics as context, never as a target. A benchmark tells you roughly where similar pages land; your own data tells you whether a change worked. Pick the one public figure that matches your page type and traffic, then build a baseline from your last 90 days and improve on it.
When you're ready to work on the page itself, our CRO guide for SaaS explains where to start when traffic is too low to test.
