A landing page converting between 2% and 5% is performing normally for most ecommerce and SaaS traffic. Below 1% something is usually wrong, and above 10% the traffic is almost always warm rather than the page being exceptional.
That range is wide enough to be nearly useless on its own, which is the problem with benchmarks. What follows is how to narrow it to a number that means something for your specific page.
The benchmark depends on the source, not the page
The same page converts at wildly different rates depending on who arrives. Ranked roughly from coldest to warmest:
- Cold paid social: 0.5% to 2%. The visitor was not looking for you and had no intent thirty seconds ago.
- Cold search on a broad term: 1% to 3%. Some intent, no familiarity.
- Search on a specific commercial term: 3% to 8%. They typed something close to what you sell.
- Email to your own list: 5% to 15%. They already agreed to hear from you.
- Retargeting: 5% to 20%. Second visit, already interested.
A 2% rate on cold paid social and a 2% rate on your own email list describe completely different situations. Compare against your own source, never against a blended industry average.
Price changes the number too
Higher prices mean longer consideration and lower immediate conversion, and that is not a fault. A $19 impulse product and a $400 considered purchase should not be held to the same rate. For the higher price the relevant metric is often not the sale at all, but whether the page produces the next step, since the decision happens over days.
Telling a page problem from a traffic problem
This is the distinction worth getting right, because rebuilding a fine page to fix bad traffic wastes weeks.
It is probably traffic when: bounce is high and time on page is very short, the rate varies enormously between sources while the page is identical, or the paid rate is terrible while organic on the same page is acceptable. People arriving with the wrong expectation leave before reading anything, so nothing on the page can be blamed.
It is probably the page when: visitors scroll the whole page and still leave, the rate is uniformly poor across every source, or people reach the checkout and abandon there. A visitor who read everything and declined was not confused about what you sell; they had an objection you did not answer.
Why the average is the wrong thing to optimise
Conversion rate is an outcome, not a lever. You cannot edit it directly. What you can edit are the specific reasons individual visitors said no, and those reasons are not visible in the rate itself.
A page at 1.5% is not failing 98.5% of the time in one uniform way. It is failing for perhaps five distinct reasons in different proportions: the price felt unjustified, the product's fit for this buyer was unclear, trust was missing, a detail was not answered, the next step was ambiguous. Fixing the largest of those moves the number; averaging them tells you nothing about which to fix.
The six recurring shapes of that objection are set out in traffic but no sales: the six objection types.
What to do before rebuilding
- Segment by source. Split the rate by channel before concluding anything. This alone resolves a large share of apparent page problems.
- Watch the drop-off point. Where people stop scrolling tells you which section lost them, and that is a specific, fixable location rather than a vague verdict.
- Get the objection in words. The rate says how many left. It never says why. Until you have the reason stated in a sentence, every change is a guess.
- Change one thing. Rebuilding everything at once means learning nothing, since a rate that moves cannot be attributed to any of the twelve changes.
Step three is the one most teams skip, because it is the only one you cannot do from an analytics dashboard. Our sample report shows what the stated objections look like when they are collected properly.
Frequently asked questions
How much traffic do I need before the rate means anything?
Rough working minimum is a few hundred visitors and at least 25 to 30 conversions before differences are worth acting on. Below that, apparent changes are usually noise, and small samples produce confident wrong conclusions.
Is a high conversion rate always good?
Not necessarily. A very high rate sometimes means traffic volume is too small and too warm, or that the price is below what the market would bear. A page converting at 25% on tiny traffic is worth investigating rather than celebrating.
Should I A/B test?
Only with enough traffic to reach significance in a reasonable time. Most small sites do not have it, and running underpowered tests for months produces confident nonsense. Qualitative feedback is the better tool below that threshold.
Does page speed affect this?
Yes, and it is worth fixing first because it is unambiguous. Slow pages lose visitors before any messaging matters. Once load time is reasonable, further speed gains stop being the constraint and the message becomes the constraint.