Conversion tracking: counting the thing that actually matters

Conversion tracking is the recording of the moment somebody does the thing a page was built to make them do — submit the enquiry, book the call, buy the item. Without it, every marketing decision is made on traffic, which measures how many people arrived rather than whether anything happened. Most small companies have analytics installed and no conversions defined, which is the same as having none.

Defining what counts as a conversion

• One primary event per page, which is the action the page exists for. More than one and nothing is primary.
• A small number of secondary events worth knowing about: a brochure downloaded, a price page reached, a phone number tapped.
• Events that are genuinely verifiable, not scroll depth or time on page, which measure patience rather than intent.
• The quality of the conversion if you can get it — an enquiry that became a customer differs from one that never replied.
• Not everything. A tracking setup with thirty events gets reviewed by nobody and maintained by nobody.

Why conversion tracking numbers never agree

Your form system, your analytics and any advertising platform will each report a different figure for the same week, and all three can be right. They count different moments — a submission received, a page view after submission, a click attributed within a window — and they attribute to different things. Ad platforms count a conversion against the click they were paid for, including view-through, which is why their figures are reliably the highest. Pick one source as the number of record, usually the one closest to the actual event, and treat the rest as directional.

Check for double counting before trusting any rise. A thank-you page that visitors can refresh, reach from a bookmark or land on twice will inflate conversions quietly, and the fix is counting the submission itself rather than the arrival at a page.

Setting it up without a platform

1. Start from the record of real submissions, which you already have, rather than from the analytics interface.
2. Define the primary event in one sentence and write down which system is the number of record.
3. Compare the two counts for a month and understand the gap before using either figure for a decision.
4. Exclude internal traffic, which in a small company can be a material share of visits to your own pages.
5. Record the source alongside the conversion, since a conversion with no attribution cannot change where you spend.
6. Review quarterly against what became revenue; an event nobody has checked against outcomes is a vanity metric with a technical dependency.

The honest baseline

If none of this is in place, the useful first step is not a tracking project. Count the enquiries you received this month and where each came from, by hand, from your own records. That number is slow, accurate and sufficient to make decisions with, and it is the figure every tracking setup should be checked against. Teams who skip it end up trusting a dashboard that disagrees with their own inbox.

Ettex Forms records each submission with its source and time, which makes the enquiry itself the number of record rather than a page view that stood in for it. Where the link that produced the visit was tagged, the campaign labels line up with the submissions rather than having to be reconciled afterwards.