Greg K.SEO and analytics at Adcore · LinkedInFour thousand people visited the site, and seventy-two enquiries came in. Whether that is good or bad is impossible to tell from those two numbers until you put them together. This article covers what a conversion and a conversion rate are, the formula for the rate, where to find your own number in analytics, and what every percentage point is worth in money.
- What a conversion is, in plain terms
- How to calculate conversion rate: the formula
- What to count as a conversion: macro and micro
- Sales conversion and website conversion are not the same thing
- Where to find your own conversion rate
- What one conversion costs
- What counts as a normal conversion rate
- Why a conversion rate is low: what to check first
- What “conversion” means outside marketing
- Where to start
What a conversion is, in plain terms
A conversion is what happens when a visitor does the thing the page exists for: sends an enquiry, buys, calls, books an appointment. The word comes from the Latin conversio, “transformation”. A random visitor is transformed into someone who has written to you or paid you.
Alongside the event itself, people almost always count its share. When someone says a site converts at two percent, they are not talking about two enquiries — they mean that out of every hundred visits, two ended in one. The correct name for that number is conversion rate, often shortened to CR. In conversation it is often called “conversion” too, but these are different things, and this article keeps them apart: a conversion is the action, the conversion rate is the share of visits that end in one.
What counts as the action is decided by the site owner, and that is the most important part of the whole definition. A conversion in marketing is not necessarily a purchase. For an online shop it is a completed order, for a services site a submitted form or a phone call, for a clinic an appointment, for a blog a subscription. The same site can count several different conversions at once.
In short. A conversion is an event. The conversion rate is its share as a percentage. Which event counts is different for every business, and that choice determines the entire number.
How to calculate conversion rate: the formula
How to calculate a conversion rate takes one line: the formula is simple enough to do in your head. Take the number of conversions, divide by the number of visits over the same period, and multiply by a hundred.
The difficulty is not the arithmetic but what you put into it. Three places where people go wrong most often:
Rather than working it out on paper, put in your own numbers. The third field is the average order value: it turns percentages into money, and from there you can see what each tenth of a percent is actually worth.
Three numbers from last month's stats — and you see what each percent is worth.
The estimate shows the order of magnitude, not a forecast. Traffic and average order value stay fixed here — only the conversion rate changes.
The slider shows the main reason anyone calculates a conversion rate at all. Going from 1.8% to 2.3% looks like a rounding error — half a percentage point. In money, at the same four thousand visits, it is twenty extra enquiries and seventy thousand hryvnia a month, without a single coin spent on additional traffic.
What to count as a conversion: macro and micro
Most sites count one event — the one that brings in money. It is called a macro conversion: an order, an enquiry, a call, a booking. There is only one of them, and it is what people mean when they ask about a site's conversion rate.
The problem is that a single number does not show where things broke down. A visitor might not have reached the service page, or reached it and never opened the form, or opened it and abandoned it halfway. Each of those steps has its own micro conversion — an intermediate action that brings no money but shows the person is moving in the right direction.
The same 1.8% broken down by step. The biggest loss did not happen at the form but earlier: three quarters of visitors never reached a service page at all.
This example shows exactly why the intermediate numbers matter. If all we saw were the final 1.8%, the logical move would be to work on the form — it is the last thing before an enquiry. But 310 people opened the form and 72 submitted it: 238 lost there. Between arriving on the site and reaching a service page, 2,800 disappear. The work belongs where the losses are bigger.
Micro conversions worth having on almost any site:
- a view of the pricing page;
- a tap on the phone number from a mobile;
- opening the booking form or the cart;
- a jump to a messenger or to the directions map;
- reading the page to the end.
None of them is a result in itself. Together, though, they turn one blank number into a route with a visible point of breakage.
Sales conversion and website conversion are not the same thing
They are two different things, and confusing them is expensive. A sales conversion is an enquiry that turns into a deal, and the sales conversion rate is the share of deals out of all enquiries that reached a salesperson. A website conversion happens earlier: a visit that turns into an enquiry. These are two adjacent stretches of the same path, and different people are responsible for them.
| Conversion rate | What we divide | By what | Who influences it |
|---|---|---|---|
| Website | enquiries | visits | marketing, the site, traffic |
| Sales | deals | enquiries | sales team, price, service |
| End to end | deals | visits | both together |
Take the numbers from the previous section. The site produced 72 enquiries from 4,000 visits — a website conversion rate of 1.8%. Of those 72, eighteen became clients — a sales conversion rate of 25%. Multiply them and you get the end-to-end figure: 0.45%, that is 18 clients from the same 4,000 visits.
Why separate them. If the website conversion rate drops while sales from enquiries hold steady, the problem is in the traffic or on the page. If the site delivers the same number of enquiries but fewer deals close, the site has nothing to do with it and changing anything there is pointless. Without the split, these two different problems look identical: “conversion is down”.
Retail tracks one more rate — the store conversion rate: the share of visitors who bought something. For a physical shop that is people leaving with a purchase out of everyone who walked in; for an online shop, orders out of visits. The logic is the same, only the definition of an entry changes.
Where to find your own conversion rate
You rarely have to count by hand: the number is already in the free Google Analytics 4 that sits on almost every site. There is one catch — GA4 does not use the word “conversion” where you would expect it. The reports say key events, and that is the same thing.
The third tile on the home screen is key events. Next to it sit 3,100 events in total: analytics sees everything that happens on the site but treats as a result only what it was shown.
The gap between those two numbers — 3,100 and 89 — is the single biggest misunderstanding for newcomers. The system really does record everything: every page view, every scroll, every file download. But it has no way of knowing which of that is a result for you and which is just movement. Until you point at the event you need, your conversion reports will read zero while the analytics is working perfectly.
The same site from the inside. Events named click_google_map and click_instagram did not appear by themselves — somebody set them up, because those are the ones that mean something for this business.
The whole path from activity on the site to a number in a report takes four steps, and the third is done by hand.
The third step takes a minute: in the GA4 admin you mark the event you need. Until you do, everything else runs idle.
Where to look for what:
- GA4, home screen — how many key events in the period and which way they are moving;
- GA4, Traffic acquisition report — the same conversion rate per source: search, ads, social. This is where you see which channel brings people and which brings passers-by;
- Google Ads — conversions from paid traffic only, right next to the spend;
- Search Console — no conversions at all, only impressions and clicks from search. Its data has to be matched to GA4 by hand.
- whether at least one event is marked as a key event — without that, the reports show no conversions;
- whether one action is being counted twice — for instance both the button click and the thank-you page;
- whether an internal traffic filter is cutting out your own visits: your own sessions and test submissions easily add spurious percentage points on small volumes.
What one conversion costs
Percentages become meaningful once you turn them into money. That is quickest to see on paid traffic, where every visit carries a price tag.
One month of an ad account. Two numbers sitting side by side — spent and received — give a third one that is not on the screen.
UAH 5,095.50 of spend against 12 conversions works out at UAH 425 per enquiry. The metric is called cost per acquisition, or CPA. On its own it is neither good nor bad: everything depends on comparing it with what a client brings in. If an average deal is UAH 3,500 and one in four enquiries closes, a client costs around UAH 1,700 in advertising — which pays off. If a deal is UAH 900, the same advertising runs at a loss.
This is where it gets interesting. There are two ways to lower the cost of an enquiry: pay less per click, or get more enquiries out of the same clicks. The first runs into the auction and your competitors. The second runs into your own site.
| Before | Conversion rate doubled | |
|---|---|---|
| Ad spend | UAH 5,095 | UAH 5,095 |
| Clicks | 234 | 234 |
| Enquiries | 12 | 24 |
| Cost per enquiry | UAH 425 | UAH 212 |
Not one extra click — the same volume, the same budget, the same auction. The only thing that changed is the page people land on. That is why raising the conversion rate is often called the cheapest way to make traffic cheaper: you are not buying more visits, you are no longer wasting the ones already paid for.
The same holds without advertising, the price tag per visit is simply less obvious. Organic traffic has a cost too — the money and months put into search engine optimization. The difference is only that nobody invoices you per click.
What counts as a normal conversion rate
This is the first question people ask once they see their own number, and the only one with no honest short answer. Figures along the lines of “2–3% is normal for a website” have circulated for decades, but they cannot be applied to a specific business: they are assembled from things that are not comparable.
Here is what changes a conversion rate several times over without changing the quality of the site at all:
So the working answer looks different. The benchmark is you, a month ago. Take your own number as the starting point and look at it in three cuts:
- Over time. Month against month, measured the same way. If the number went up after changes to the site, the changes worked — regardless of any market average.
- By source. Organic, paid, social, direct — separately. A channel converting three times worse than its neighbours is either bringing the wrong people or sending them to the wrong page.
- By page. Two similar service pages with different conversion rates is the most useful finding there is: you can see what the weaker page lost, and apply the fix everywhere else.
The only external comparison that genuinely means anything is against your own competitors for the same queries. But you will never see their internal numbers, so in practice the first method is what remains.
Why a conversion rate is low: what to check first
When the number disappoints, people usually start with buttons and colours. That is the least productive place of all. Here is the order worth working through — from the most common causes to the rarest.
- The wrong people are arrivingThe most common cause, and it is not on the site. If a page about designing houses attracts searches for “how to build a house myself”, the conversion rate will be low whatever the design. Check it in the source and query reports: look at what people arrive with.
- The page does not keep the promiseSomeone clicked an ad that showed a price, and there is no price on the page. Or the search said “repairs in two days” and the site says nothing about timing. A gap between expectation and page kills conversions faster than any technical fault.
- No answer to “how much”Missing prices are the most frequent reason people leave service pages. If an exact figure is impossible, a range works, or a worked example, or at the very least the point at which the work starts.
- Getting in touch is hardAn eight-field form, a mandatory patronymic, a captcha, a button below the second screen. Every unnecessary field is people who started filling it in and gave up.
- The site is slow on a phoneMost traffic is mobile. A page that takes four seconds to open loses some of its visitors before they have seen anything at all.
- No grounds for trustNo reviews, no examples of work, no names, no address. For services this often weighs more than price.
- The analytics is lyingAnd only after all of the above is it worth doubting the number itself: double counting, unfiltered internal visits, a conversion nobody re-marked after a move to a new version of the site.
A working rule: first find where the most people are lost — using the section on micro conversions above — and work on that step. Improving a button at a step that 95% of people already pass will not change the outcome.
What “conversion” means outside marketing
The word carries several meanings, which is worth remembering when you look it up in a reference work or read somebody else's text. They all descend from the same idea of transformation, but the fields differ.
| Field | What it means |
|---|---|
| Digital marketing | A target action by a site visitor: an enquiry, a purchase, a call. The share of visits that end in one is the conversion rate — the subject of this article |
| Finance | Exchanging one currency for another, or restructuring a debt on new terms |
| Economics and industry | Switching production from military output to civilian |
| Chemistry | The share of a starting substance that entered into the reaction |
| Linguistics | Forming a new word by moving it to another part of speech without affixes |
| Medicine and psychology | Internal tension expressed as physical symptoms |
In marketing conversations the meaning is always the first one. The rest turn up in texts from entirely different fields, and they are the reason dictionary definitions of the word often have nothing to do with websites.
Where to start
If you do not currently know your own conversion rate, here is the order of work for the coming week.
- Name one main action. The one after which you have an enquiry or money. One, not five.
- Check that it is marked as a key event in GA4. A minute of work, without which every later number will be zero.
- Add two or three micro conversions. A pricing view, a tap on the phone number, opening the form — so you see the route, not just the outcome.
- Calculate the current number and write it down. That is the baseline you will compare everything against.
- Look at the conversion rate by source. More often than not, the channel bringing the wrong people is visible right there.
- Work on the step with the biggest losses. Not the one that is easiest to fix.
One last thing, about orders of magnitude. Raising the conversion rate pays off most when traffic is already arriving: half a percentage point on four thousand visits is twenty enquiries, on two hundred visits it is one. If there are still few people on the site, it makes sense to deal with their number first — what that work involves and what it costs is covered in the articles on search engine optimization and the cost of promotion. If traffic is there and enquiries are not, the conversion rate is exactly what to work on — and that is usually both cheaper and faster.
We can go through a specific page together and find the point where it breaks — that is part of what our agency does.