Conversion tracking that counts enquiries rather than page views

On paper, before anything technical.

For most businesses the list is short: the enquiry form, the phone call, the quote request, the booking, and possibly a chat conversation that reaches a real person. Four or five items.

For each one, ask what it is worth and whether it could become revenue. If the answer is unclear, it does not belong on the list.

Things that are not leads, and are counted as leads constantly: newsletter signups, PDF downloads, video plays, time on page, scroll depth, visits to the contact page. All of these can be worth measuring. None of them is somebody asking you to sell them something, and optimising a campaign toward them produces traffic that downloads things and never calls.

The default setup counts a view of the thank-you page. It is easy and it is wrong often enough to matter.

Three ways it breaks. Somebody reaches the page from the navigation or an old bookmark, and gets counted. Somebody refreshes after submitting, and gets counted twice. The site is a single page application, and the view fires on a route change whether or not the submission succeeded.

Where your form can fire an event on successful submission, use that. The event fires when the thing you care about actually happened, and only then.

If a thank-you page is your only option, at minimum make sure it is not reachable from the navigation, and set it to noindex so it does not collect organic traffic.

For most service businesses this is the biggest hole in the numbers.

The proper version is dynamic number insertion: a tracking number displayed on the site that swaps depending on the source, so the call is attributable. It costs money and it works.

The cheap version is tracking clicks on the tel link. This is not a call. Somebody can click and hang up, and calls from a desktop where the number is copied out by hand are invisible. It is still far better than counting nothing.

Whichever you choose, be clear about what it is. A business that treats tel link clicks as calls will overstate its lead count. A business that ignores the phone entirely will understate it and misjudge which channels work, because the channels that drive calls will look dead.

Define each conversion once, then bring it where it needs to go.

The usual sensible arrangement is to define the events in GA4 and import them into Google Ads as conversion actions. One definition, one source of truth, and if the definition changes it changes everywhere.

The failure mode is defining the same action twice: once as a GA4 event imported into Ads, and again as a native Google Ads tag on the same form. Now every enquiry is counted twice, the cost per conversion looks half what it is, and bidding is being told the account is twice as successful as it is.

If you inherit an account, check for this specifically. It is common and it is invisible unless you go looking.

Google Ads lets you mark conversion actions as primary or secondary, and GA4 has key events. Use them.

Primary means it influences bidding. Only actions on your short lead list belong here. If a newsletter signup is primary, Smart Bidding will find people who sign up for newsletters, and it will be very good at it.

Secondary means it is recorded and reported but does not steer anything. This is where downloads, signups and softer engagement belong. Useful for understanding, not for optimising.

This is the step everybody skips and it is the one that finds the problems.

Take a full calendar month. Count the enquiries you actually received: emails in the inbox, calls in the log, forms in the CRM. Deduplicate the people who filled in a form and then rang.

Now compare with what the tracking reports.

If tracking is higher, you are double counting somewhere, or counting something that is not an enquiry, or a bot is filling in your form. If it is lower, something is not firing, or consent is blocking it, or a channel is not tagged.

Being within a reasonable margin is fine. Being out by a factor is a setup problem, and until it is resolved every optimisation decision made from the data is a guess.

Tracking fails silently and it fails often. A theme update, a form plugin change, a new consent banner, a developer removing a script that looked unused.

The site keeps working. The forms keep arriving. The number just stops, and nobody notices for six weeks because a downward trend looks like a bad quarter.

Two cheap protections. Set an alert for conversions falling below a floor over a few days. And put the reconciliation into your monthly routine: real enquiries counted by hand against reported conversions, one line in the report. It takes ten minutes and it is the only thing that reliably catches this.