What is conversion tracking? A plain explanation.

Conversion tracking is the practice of recording when a visitor completes an action you care about, such as submitting a form, calling a number or buying something, and connecting that action to the campaign that brought them. It answers a single question: did the marketing produce the outcome. For software that keeps the lead behind the count, see conversion tracking software.

Here is the sentence this whole post hangs on: Conversion tracking tells you an event happened. A lead source tells you who it was and where they came from. Everything below is a longer way of saying that, plus the practical part about getting the first one right.

Why conversion tracking matters

Marketing runs on a handful of channels, and each channel asks for budget. Search ads, social ads, email, content, a partner, a directory listing. Every one of them is supposed to produce enquiries, and every one of them has a dashboard that says it did.

Conversion tracking is what turns those dashboards from opinions into counts. Without it, the only evidence a campaign works is the click, and a click is a visit, not an outcome. With it, you can say that this campaign produced this many form submissions last month, and that number can be compared with what the campaign cost.

It also feeds the machinery. Ad platforms use conversions to decide who to show your ads to and how much to bid. A platform that is told nothing optimises for clicks. A platform that is told which clicks became enquiries optimises for enquiries. That alone is a reason to get it set up properly.

Common types of conversions

A conversion is whatever you decide is worth counting. Most businesses end up with some mix of these:

Form submission

A visitor fills in a quote request, a contact form or a booking form. For most service businesses this is the main one.

Phone call

A visitor calls the number on your site or in your ad. Counted by the call itself, usually past a minimum length.

Purchase

A visitor completes a checkout. Each sale is a separate conversion, and it can carry a value.

Sign-up

A visitor creates an account or starts a trial.

Chat or message

A visitor starts a conversation through a chat widget or a messaging app.

Offline outcome

A lead becomes a customer later, outside the website. The sale happens in a CRM or on a phone call and is sent back afterwards.

The type matters less than the discipline. Pick the handful that map to money, and ignore the ones that only look busy.

Conversion tracking examples

Three illustrative setups show the idea.

A dentist runs a search campaign for emergency appointments. The booking form is the conversion. The tag fires when the booking is confirmed, and the campaign report shows bookings, not clicks.

An accountancy firm sends a monthly newsletter. Nobody buys from an email, so the conversion is a consultation request from the link in it. The count shows whether the newsletter produces enquiries or only opens.

A software company offers a free trial. The conversion is account creation, and a second one, a paid upgrade, is recorded later. Two conversions describe two stages, and each has its own count.

In every case the setup is the same: a defined action, a tag that fires once when it succeeds, and a report that attributes it to a campaign. The setup steps for one platform are in how to set up Google Ads conversion tracking.

Conversion rate, and what it hides

Conversion tracking produces two numbers. The count is how many conversions happened. The conversion rate is that count divided by the number of visitors, usually shown as a percentage. If a page gets a thousand visits and produces twenty enquiries, its conversion rate is two percent.

The rate is useful for comparing pages. It tells you which landing page turns visitors into enquiries more often, and it is the figure most testing is built on. What it hides is the quality of the enquiry. Two pages can share a rate and still produce very different customers, because the rate treats every submission as equal. A form that attracts window shoppers and a form that attracts people with a budget look identical on the chart.

That is why a conversion count is a starting point. It tells you where to look. Deciding where the money goes needs the next layer: who the enquiries were, and which of them turned into customers.

Common mistakes when starting out

• Counting everything. Every click and scroll becomes a conversion, the report fills up, and the ones that matter are lost in the noise.
• Counting the same outcome twice. One form submission reaches the platform through a tag and through an import, and a single lead is reported as two.
• Trusting a number nobody tested. The tag was added once and never checked. It could be firing on the wrong page, or not at all.
• Comparing tools that count differently. One platform says forty, another says thirty-one, and the afternoon goes on deciding which is lying. Neither is.
• Stopping at the count. The report says the campaign converted. It cannot say who converted, so nobody knows whether to fund it again.

How ad platforms and analytics record conversions, and why they disagree

Each tool keeps its own tally, and it uses its own rules to do it. An ad platform counts the outcomes it can connect to its own ads. An analytics tool counts every occurrence of the event, from whatever source. Add consent choices, blocked scripts and browser privacy controls removing data in different places, and the totals drift apart.

This is normal, and it surprises people every time. Two tools reporting forty and thirty-one for the same campaign are usually both working. They are answering slightly different questions. A walkthrough of how Google's analytics product handles it is in GA4 conversion tracking: what key events actually tell you.

There is a second, quieter limit. A count has no name attached. You learn that thirty-one people converted, and nothing about who they were, which of them were worth having, or which channel they really came from when the ad platform only credits its own. That is the gap lead tracking fills, and it is why lead source categories exist at all.

How to track conversions accurately

Accurate conversion tracking has four requirements.

1. Define the outcome. Write down what counts, in words, before touching a tag. "A completed quote request" is a definition. "Engagement" is a mood.

2. Fire once, on success. The tag should run when the submission succeeds, not when the page loads or the button is clicked. A tag on the wrong moment counts attempts and refreshes.

3. Use one path per outcome. If the same event reaches a platform by two routes, it is counted twice. Pick one route for each conversion and keep the other off.

4. Check it. Submit a test, watch the tag fire, and look at the status inside the platform. Check again after every change to the site. A setup that was right in March is not guaranteed in October.

When the outcome happens away from the website, such as a sale that closes on the phone, the same discipline applies, with one extra step: the identifiers have to be saved when the lead arrives, so the outcome can be sent back later. Offline conversion tracking covers how that works today.

For the full list of values to standardise on, see the lead source taxonomy.

Keep the count and the person

Conversion tracking is necessary, and it is only half of the picture. It gives you a number you can put in a report. The other half is the record of who converted and where they came from, which is what lets you act on the number.

Lead Source records the source channel, the campaign, the landing page and the full page journey behind every form submission, with first touch and last touch both kept, and names the person when the form captures a name. Keep your conversion tracking as it is. The conversion tracking page shows how the two work side by side, and what a lead source is defines the other half in the same plain terms.