Why Two Stores With the Same CVR Aren't Equal

Two stores with the same conversion rate are rarely in the same situation, and the reason is structural rather than statistical. A conversion rate is orders divided by sessions, and almost everything that determines the answer sits outside the thing the number is used to judge. Last updated: September 2026.

Omniconvert has measured conversion across the CROBenchmark dataset of 7,000+ websites in 15+ industries, against 248+ audit criteria, over 13 years in eCommerce. The pattern that shows up again and again is two stores reporting nearly identical rates where one is a well-built site fighting for cold demand and the other is an ordinary site harvesting a strong brand. The rate cannot tell those apart, and teams spend quarters acting as though it can.

This article is about what sits underneath the number. For where the level ought to be in the first place, see a good conversion rate.

Related reading: metrics that predict conversion, beyond CVR.

Why two stores with the same conversion rate are not equal

Start with the denominator, because it is the half people forget. Sessions are not a fixed quantity of opportunity. A store buying broad awareness traffic and a store fed by people searching its name are counting two different kinds of visit, and only one of those groups arrived intending to buy.

The numerator is shaped by the category more than by the site. Something cheap, familiar and reordered regularly converts at a rate a considered purchase will never reach, because the decision was already made before the session started.

Put those together and an identical rate becomes almost uninformative across two businesses. It is a single number describing the interaction of a market, a catalogue, a media mix and a website, and it is routinely read as a score for the last of those four.

This is not an argument for ignoring the rate. It is an argument for never comparing a blended one.

The six variables underneath the number

Branded traffic share. Visitors who searched for you by name convert at a multiple of those who found you another way, because they already chose. A store with strong brand demand carries a high blended rate regardless of how its pages perform, and the site's actual weaknesses are only visible in the non-brand segment.

Device mix. Phones convert lower than desktops in most categories, so a store with a heavier mobile share posts a lower blended rate while potentially outperforming on each device individually. This is the single easiest correction to make and it is skipped constantly.

Returning-customer share. Returning buyers convert far better than new ones. A mature store with a strong base can look excellent in blended terms while acquiring poorly, which is a serious problem hiding behind a good number.

Average order value. Expensive purchases involve more research and more visits, so they convert lower by construction. A low rate on a high-value catalogue often describes a healthier business than a high rate on cheap goods.

Purchase frequency. Categories people buy weekly behave differently from categories people buy once every few years, both in how often somebody returns and in how much of the deciding happened before arrival.

Session definition. The least interesting variable and one of the most powerful. Timeout settings, bot filtering, consent handling and whether the conversion event fires at payment or at confirmation all change the figure by meaningful amounts.

Measurement differences come first

Bot and crawler traffic is the most common single distortion, and it moves in one direction: it inflates sessions and depresses the rate. A store filtering aggressively and a store filtering loosely are not measuring the same denominator.

Consent handling changes the picture too. Where analytics only records visitors who accepted tracking, both halves of the ratio are affected, and stores in different markets with different consent behaviour are not comparable on this number at all.

Then there is where the event fires. A conversion recorded when the payment button is pressed counts attempts including failures. One recorded on the confirmation page counts completions. The gap between those two definitions is the size of your payment failure rate, and it is invisible unless somebody checks.

None of this is exotic, and all of it is checkable in about an hour. Do it before any conversation about why a competitor's published rate is higher than yours.

What the same rate can mean in practice

The discounting row is the one that produces the worst decisions. A conversion rate bought with a permanent promotion looks like an optimisation success and reads as a margin problem only much later, in a different report, owned by a different person. Revenue per session catches it immediately, which is the argument for tracking that number beside the rate rather than instead of it.

How to compare two rates honestly

The practical version is short. Split your own rate by device and by source, and hold the non-brand segments separately, because that is where the site's own contribution is visible. Then compare those segments over time against themselves.

External figures are harder, because published benchmarks are almost always blended and rarely say what they contain. Use them to establish whether you are in a plausible range, and never to set a target.

When a competitor's number does become available, the useful question is not whether theirs is higher. It is which of the six variables differs, because every one of them suggests a different response, and several of them suggest doing nothing at all.

Baymard Institute's checkout research puts average cart abandonment near seventy percent across the industry, which is a helpful reminder of the scale of the shared problem: almost every store in every one of these pairs is losing most of its carts, and the differences between them are smaller than the thing they have in common.

What the rate can never tell you

Returns are the clearest example. A change that raises conversion and raises returns by more has cost money, and the conversion report will show a win for as long as anybody looks only at the conversion report.

Acquisition cost is the second. A rate that improved because the media mix narrowed to the cheapest, most ready audience is not an improvement in the site, and it usually comes with a ceiling that arrives a quarter later.

And then there is what happens after the order. Bain and Company's work with Fred Reichheld holds that a five percent improvement in retention can raise profits by twenty-five to ninety-five percent, which is a far larger lever than most conversion work, and it is entirely invisible in this metric. To find out what is actually costing you orders rather than what your rate is, audit it.

What to do this week

• Check the definition first. Bot filtering, session timeout, where the conversion event fires. An hour, and it occasionally makes the rest of the exercise unnecessary.
• Split the rate four ways. Brand and non-brand, mobile and desktop. Four numbers instead of one, and the interesting one is non-brand mobile.
• Put revenue per session beside it. The one number that notices when a conversion gain was bought with discount.
• Retire the blended figure from decisions. Keep reporting it if the board expects it. Stop letting it start projects.
• Validate any change against a control. Omniconvert Explore is the CRO platform for that: A/B and multivariate testing, on-site surveys, heatmaps and session recordings. Where the constraint is that order, return and session data live apart, Nexus by Omniconvert unifies that data and ranks the next actions by True Profit, with a human approving what goes live.

FAQ: comparing conversion rates between stores

Why can two stores have the same conversion rate but very different results?

Because a conversion rate is a ratio, and both halves of it can change for reasons that have nothing to do with how good the store is. Traffic composition, device mix, brand versus non-brand demand, average order value, purchase frequency and how sessions are counted all move the number. Two stores can arrive at an identical figure from opposite directions, and the more useful comparison is what each one is made of rather than where each one lands.

Which variable distorts conversion rate comparisons most?

The share of branded traffic. Visitors who searched for you by name arrive with the decision largely made, and they convert at multiples of the rate of people discovering you. A store with strong brand demand can carry a mediocre site and a healthy conversion rate at the same time, which is why the number alone never tells you whether the experience is working.

Does average order value affect conversion rate?

Strongly, and in the direction people forget. Expensive and considered purchases convert at lower rates because the decision takes longer and involves more visits, so a low rate on a high-value catalogue can describe a healthier business than a high rate on cheap impulse goods. Comparing the two rates directly is comparing two different buying behaviours.

How should I compare my conversion rate with another store?

Segment before you compare. Split by device and by source, and compare like with like: your non-brand mobile rate against theirs, not your blended figure against their blended figure. If you cannot obtain their segments, which is usually the case, treat any external rate as a rough orientation and compare against your own history instead.

Can analytics differences explain a gap in conversion rate?

Frequently, and it is the first thing to rule out. Session timeouts, bot filtering, cross-device handling, consent-banner effects on measurement and whether the conversion event fires before or after payment confirmation all shift the number. A gap of a third between two stores can be entirely a measurement difference, and that check costs an hour.

What should I track instead of conversion rate alone?

Keep the rate, add composition and add value. Conversion rate split by device and source, average order value, revenue per session and repeat purchase rate together describe a business in a way that a single rate cannot. Revenue per session in particular catches the case where a rising conversion rate is being bought with discounting.