A website report can show more leads than the office received. It can also miss requests that reached the business successfully. Before using those numbers to change an advertising budget, test what the website actually counts and compare it with the requests your team can find.

For a local service business, a useful tracking audit follows one customer journey from the first interaction to a recorded outcome. Start with the form, phone link, and booking process customers use most often. The checks below help an owner and website provider distinguish a measurement problem from a problem handling inquiries.

Counting a button click as a received inquiry

A visitor can click Submit without completing a valid request. A required field may be empty, the connection may fail, or the website may reject the submission. A tracking rule attached only to the button can count that attempt even though the office has nothing to respond to.

Ask what event your lead total represents. Google's enhanced measurement documentation distinguishes form starts from form submissions. Neither label, on its own, proves that a particular website's form integration delivered a usable inquiry to your team.

Test an incomplete submission, then a valid one. Check the website confirmation, the saved form record, and the destination inbox or customer system. The event used to count received inquiries should correspond to successful receipt. Keep unsuccessful attempts separately if they help diagnose form problems.

Counting the same action more than once

A site can have several ways to send tracking data: a plugin, custom code, or a tag manager. Two implementations may track the same successful submission. A confirmation-page rule can also count another event when someone refreshes or revisits the page, depending on its setup.

Submit one clearly labeled test request and ask your provider to inspect the relevant events. Check whether the same lead event appears twice and identify which implementation sent each occurrence. Do not remove tracking blindly; different destinations may legitimately need their own events.

There is another duplication problem to check in the report itself. If form submission and confirmation-page arrival describe the same request, adding both to a combined lead total overstates demand. Document the events included in that total, then choose one reliable counting method for each type of inquiry.

Treating a phone tap as a customer conversation

Tapping a phone link can open a dialer without producing a completed call. Even a connected call may concern an existing appointment, a supplier inquiry, or something outside your service area. These actions need different labels.

Keep phone-link clicks, received calls, and qualified service inquiries separate where the available systems allow it. If your setup measures only taps, name the metric “phone-link clicks.” Do not present it as calls answered or jobs generated.

Test the displayed number and clickable number on a mobile device. Where a tracking number replaces the normal number, check that calls reach the correct destination. Then review how the office records the outcome. Call duration can help with screening, but a long conversation alone does not establish that someone booked.

Losing the request at an external booking page

A visitor may leave your website to use a scheduling service. If you track only the link that opens that service, you know someone began the journey. You do not yet know whether an appointment was confirmed.

Complete a test booking through the same route a customer would take. Check what is recorded on the website, in the scheduling platform, and in the calendar. Ask your provider whether the scheduling tool supports an appropriate confirmation event or integration before promising complete measurement.

If you cannot connect the systems reliably, report booking-page visits and confirmed appointments separately. State the limitation. Also test cancellations and rescheduling so the records distinguish a newly booked appointment from a change to an existing one.

Giving one customer several lead records

Imagine a homeowner submits a request about a heating problem, calls to check that it arrived, and then books an appointment. This hypothetical journey contains several interactions but only one service need. Counting each interaction as a separate lead creates a misleading picture of demand.

Keep the website's interaction counts, but reconcile them with unique inquiry records in your customer system. Link repeat contacts to the original request where you can establish that they concern the same need. A returning customer asking for different work should remain a separate opportunity.

For example, six submissions and four calls might describe seven unique service requests. They might also describe ten. Inspect the records before choosing a total; the event counts cannot settle that question by themselves.

Comparing reports that answer different questions

Your website report may count submission events, your office log may count people, and your calendar may count appointments. A difference between those totals does not automatically indicate broken tracking. First establish the counting unit and period for each report.

Use the same date range and check time-zone settings when reconciling activity. Allow for reporting delays. Record whether test submissions, spam, and duplicate contacts are included. Browser settings and consent choices can also limit what appears in analytics, so the operational record remains important when checking received requests.

Finally, separate the date a request arrived from the date it booked. An appointment confirmed this week may have started with last week's inquiry. Follow a group of requests through to their outcomes when calculating a booking rate, rather than dividing unrelated weekly totals.

Retesting after website changes

A new form, redesigned landing page, or replaced booking tool can change the behavior your tracking depends on. Add measurement checks to the website release process instead of waiting for a surprising monthly report.

Maintain a short test record with the page, action, expected result, observed result, and person responsible for any fix. Repeat the successful and unsuccessful form tests, the phone-link check, and the booking journey after relevant changes. Use clearly labeled test details and remove test records from business reporting according to your team's process.

Once those checks pass, compare a small sample of real inquiries with the operational records. Document any remaining gaps and who will investigate them. You can then discuss advertising performance with a clearer understanding of what the numbers measure and which customer outcomes still need to be connected.

About the author

Guy is the owner of ScheduledHVAC, an HVAC marketing business.