Someone tells you their conversion rate is 2 percent. You have learned almost nothing. Two percent of what, measured over what window, from which traffic source, counting which action as a conversion? A 2 percent checkout completion rate would be a catastrophe. A 2 percent rate from cold social impressions to a paid customer would be extraordinary. The number is identical. The meaning is opposite.
This is why hunting for conversion rate benchmarks online is usually a waste of an afternoon. You find a figure, you compare yourself to it, and you feel either falsely reassured or needlessly panicked, because the study behind the figure measured a different step in a different funnel for a different kind of buyer. The useful move is to stop looking for someone else’s number and start building your own.
A conversion rate is a fraction, and both halves need defining
Every conversion rate is conversions divided by opportunities. Most arguments about whether a rate is good are actually arguments about the denominator. If you count impressions, your rate will look tiny. If you count people who reached the pricing page, it will look large. Neither is wrong. They are answers to different questions.
Before you record a single number, write down in plain language what you are measuring. Not "conversion rate" but "of the people who clicked a link in a LinkedIn post in September, what share started a trial within seven days." That sentence contains a numerator, a denominator, a source and a time window. If you cannot write that sentence, you cannot interpret the number that comes out.
The time window matters more than people expect. A trial-to-paid rate measured on day 3 of a 14 day trial is not a conversion rate, it is a partial count. Cohorts fix this: only measure a group once every member of it has had the full window to convert.
The steps worth measuring separately
Blended, end-to-end conversion is the number you report to yourself at the end of the month. It is nearly useless for deciding what to fix, because a change anywhere in the chain moves it. Break the chain into steps that fail for different reasons, and measure each one on its own.
- Impression to click. Fails when the hook is weak or the audience is wrong.
- Click to page view. Fails for technical reasons: slow load, broken redirect, a link preview that eats the tap.
- Page view to signup or email capture. Fails when the offer is unclear or the form asks too much.
- Signup to activation, meaning the first moment the product delivers something. Fails on onboarding.
- Activation to paid. Fails on pricing, on value, or on a checkout that asks for a card too early.
- First payment to second payment. Fails on the product itself, and it is the step almost nobody measures.
Each of these has a different owner and a different fix. Lumping them together means you spend three weeks rewriting headlines when the real problem was that your link opened inside an in-app browser that logged people out.
Why the benchmarks people quote should be treated as rumour
You will see confident ranges quoted everywhere. A few percent of impressions click. A fifth to a third of landing page visitors give an email. Some fraction of free trials become paid. These figures get repeated so often they feel like physics. They are not. Most trace back to one vendor summarising its own customer base, which is a self-selected sample of companies that bought that vendor’s product, in one industry, at one price point, in one year.
That does not make them worthless. It makes them a smell test rather than a target. If your landing page converts visitors to signups at a rate far below anything you have ever seen quoted, something is probably broken and worth a look. If you are somewhere in the vicinity, the benchmark has told you nothing you can act on, and the only useful comparison left is against yourself last month.
Be especially wary of any benchmark that does not state its traffic source. Traffic from a branded search, from an email to existing subscribers, and from a cold video on a short-form feed convert at wildly different rates for the same page. Mixing them produces an average that describes nobody.
How to build your own benchmark in about a week
You do not need a data team for this. You need one tracked link per post and a way to see which of those clicks eventually turned into money. The mechanics are boring and they work.
- Put a unique tracked link on every post that has a destination. One link per post, not one per campaign.
- Record the impression count for each post from the platform’s own analytics, and accept that these numbers are approximate and defined differently on every network.
- Log signups and payments with the source that produced them, so you can join the two ends of the funnel.
- Wait for a full conversion window to pass before you compute anything. Two weeks is a reasonable default for a low-priced product.
- Compute each step separately, per channel, and write the four-part sentence next to it.
After a month you will have something no published benchmark can give you: the actual rate at which your audience, reading your words, buys your product. That is your baseline. Every future number gets compared to it. This is the loop seenpaid runs by default, tying a tracked link on each post to the Stripe payments that follow, so the join between click and revenue happens without a spreadsheet.
When a difference is real and when it is noise
The most common analytics mistake among small teams is treating tiny samples as signal. If a post got 40 clicks and produced one sale, your click-to-sale rate is 2.5 percent, but a single extra sale would have made it 5 percent. Nothing about the post changed. The number doubled on one purchase.
A rough guard: if adding or removing one conversion changes your headline number by more than a fifth, you do not have enough data to draw a conclusion. Either wait, or pool similar posts together and compare groups instead of individuals. Comparing "posts where I told a customer story" against "posts where I announced a feature" across twenty posts each is far more informative than ranking twenty individual posts.
Watch the trend rather than the level. A step that moves from a low rate to a slightly less low rate across three consecutive months is a real improvement, even if it never approaches a number you read in a blog post.
Segment before you conclude anything
A flat overall rate often hides two opposite movements. Mobile got worse while desktop got better. One platform started sending you a lot of low-intent traffic and dragged the blended average down while absolute revenue rose. Always look at the rate split by traffic source and by device before you decide what to fix.
The same applies to your own content. Educational posts and promotional posts should not be measured against the same click rate. A teaching post that gets few clicks may be doing the trust-building work that makes a later promotional post convert. Judge each type against its own history.
The rate that actually matters
If you only track one thing, track the rate at which a person who first touched your content becomes a paying customer, measured per channel, over a fixed window. Everything upstream is a means to that. It is also the only rate that translates directly into a decision about where to spend next week.
Conversion benchmarking is not a search for a magic number. It is the discipline of defining a step precisely, measuring it consistently, and letting the trend tell you where the leak is. Tools help with the plumbing. seenpaid exists because joining a post to the revenue it produced is the tedious part, and once that join is automatic, the benchmark question answers itself: your benchmark is you, last month.