How do you calculate conversion rate?
Conversion rate is found by dividing the number of conversions by the number of visits and multiplying by a hundred. If three out of a hundred visits buy, your conversion rate is three percent. What really determines the result is not the formula but what you put in the denominator: sessions, users and ad clicks give three different answers.
The formula is simple and everyone knows it. The problem is that three different conversion rates can be calculated for the same site in the same period, and all three are technically correct.
That makes conversion rate a matter of definition rather than a single number. Any comparison made without fixing the definition is misleading.
The formula and three different denominators
You can put sessions, users or clicks in the denominator. The three answer different questions.
Denominator | What it calculates | When to use it |
|---|---|---|
Session | The likelihood that any visit converts | When measuring site performance |
User | The likelihood that any person converts | In long decision cycles |
Click | The conversion rate of an ad click | When measuring advertising efficiency |
On the same site the user-based rate always comes out higher than the session-based rate, because a person visits several times before buying. Comparing the two figures is meaningless.
What you count as a conversion matters just as much
Purchase is not the only kind of conversion. Form submissions, phone calls, add to cart and newsletter signups can all be defined as conversions.
The danger is mixing them into the same rate. If you put add to cart and purchase in the same conversion bucket the rate goes up while nothing has improved.
Measure the primary conversion on its own, usually a purchase or a qualified form
Measure intermediate steps separately, they are what show where the funnel narrows
Write the definition down, so you can be sure you are measuring the same thing six months later
What is a good conversion rate
There is no answer to this from outside. The figures circulating as industry averages were calculated with different definitions, different traffic mixes and different price ranges.
The only meaningful comparison is against your own history. What was your rate last month, on the same channel, in the same campaign.
The second meaningful comparison is by channel. Traffic from a brand search always converts better than traffic from an ad the person does not recognize. Averaging those two rates together makes both meaningless.
If your overall conversion rate is falling while your number of sales is rising, do not panic. The reason is usually more, and colder, traffic entering the top of the funnel. That is not a bad development.
Where to look when the rate drops
First verify the measurement itself, is the conversion tag still working
Look at the channel breakdown, is the drop across every channel or just one
Has the traffic mix changed, is a new campaign bringing in cold traffic
Look at the device breakdown, if the drop concentrates on mobile the cause is usually page speed
Check page speed and error rate, a page that slows down loses conversions quietly
This order matters. Any interpretation made before the measurement is verified is a guess, and a meaningful share of sudden drops in conversion rate are not a real drop at all but a broken tag.
Sources
Frequently asked questions
What is the conversion rate formula?
Should I calculate session-based or user-based?
Why do my analytics tool and ad platform show different rates?
What is the average conversion rate?
Can sales rise while conversion rate falls?
MORE ARTICLES