A campaign can look successful in an ad dashboard while the business wonders where the customers are. Clicks are coming in. The cost per conversion looks reasonable. But when someone checks the inbox, the enquiries are spam, outside the service area or asking for work the company doesn’t do.

That’s often a measurement problem as much as an advertising problem. Before spending more, I’d want to know what we’re calling a conversion and whether the tracking reflects the way the business actually handles leads.

Start by deciding what a useful enquiry looks like

A plumber, restaurant and professional services firm shouldn’t all use the same definition. One business might value a scheduled consultation; another might get most serious enquiries by phone. Write down the action that shows genuine buying interest.

For example, a local contractor could treat a completed estimate request as a primary conversion. A phone-number click might be a secondary signal because it doesn’t prove someone made a call. A confirmed booking is a later-stage business result. Those are three different things, even if the platform can report all three.

Get agreement on definitions before touching tags. Otherwise every dashboard discussion becomes an argument about what the numbers mean.

Map the customer journey before installing anything

Draw a straightforward path: someone sees an ad, visits a relevant service page, reads enough to decide and submits a form or calls. Then someone on the team responds, confirms the job is a fit and moves the enquiry toward a booking.

For each point, note what can be measured reliably. Page visits explain traffic. Form submissions demonstrate a completed action. Call tracking, where implemented appropriately, can help connect calls to campaigns. Your CRM, inbox or booking records tell you whether the lead was actually useful.

A small measurement table can include: action, definition, tool receiving the event, whether it is primary or secondary, and who verifies quality.

Track forms without counting the same lead twice

A common mistake is firing a conversion when someone clicks the submit button instead of when a valid submission succeeds. That can include people who never completed the form.

Prefer a confirmed success event tied to a successful submission. If the form redirects to a thank-you page, check that reloading it doesn’t create misleading extra conversions. If you use a single-page form, check whether both the form software and your tag manager are sending the same event.

Test a successful form and a failed validation attempt. Look at the receiving inbox or CRM too. One real test submission should produce the expected record and, where appropriate, one conversion event.

Be careful interpreting calls

A phone link click is not the same as a completed conversation. Someone may tap by accident, hang up immediately or call about something unrelated.

If calls matter, decide whether you can measure real calls, duration or qualification without collecting more personal information than necessary. A short call can still be valuable in some industries, so avoid treating a universal duration threshold as proof of quality.

The business should ideally record the outcome: in-area and relevant, out of area, existing customer, wrong service or booked job. That extra context makes ad decisions less dependent on a platform’s assumptions.

Connect Google Ads and Meta Ads to business outcomes

Both platforms can optimize using the signals you send them, but they have different attribution windows, reporting methods and privacy constraints. Don’t add together their reported conversions and assume that number equals unique customers.

Use consistent definitions across systems where possible. Make sure the event you mark as a primary optimization goal is genuinely important to the business.

Before increasing spend, compare ad reports with form records, calls and bookings over the same period. Differences deserve investigation; they aren’t automatically evidence that someone configured tracking incorrectly.

For campaign planning alongside measurement, see Google Ads for local services and Meta Ads for service businesses.

A form may collect a name, phone number, address or details about a project. That doesn’t mean those values belong in analytics event parameters.

Avoid sending personal or sensitive data into advertising tools unless there is a clearly supported, compliant workflow. Check consent behaviour and the site’s published privacy information against what the tags actually do. Platform requirements and local laws can vary.

Test with the appropriate consent states. A tool reporting no events under one state may be functioning as intended rather than being broken.

Build a ten-minute monthly check

Open your reporting and answer four questions. Are the events firing as expected? Do total form submissions broadly reconcile with the business’s own records? Are qualified leads coming from the services and areas you intended? Has a website or tag change altered the numbers unexpectedly?

If an event count doubles overnight after a redesign, investigate tags and form behaviour before celebrating. If clicks rise but qualified leads fall, review search terms, targeting and the landing page rather than automatically adding budget.

A useful conversion report shouldn’t just tell you what people clicked. It should help the business decide what to improve next.

To make the rest of the measurement setup work, review conversion tracking for local ads alongside the campaign itself.

Frequently asked questions

What is the best conversion to track for a local business?

The one that reliably represents a meaningful customer action. For many service businesses that is a completed enquiry or booking, supported by lead-quality records.

Should I count phone clicks as leads?

You can measure them as useful interaction signals, but don’t confuse them with verified calls or qualified opportunities.

Why do Meta, Google Ads and analytics show different numbers?

Attribution models, reporting windows, consent, deduplication and event definitions can differ. Reconcile against the underlying business records rather than expecting identical dashboard totals.

How do I know if tracking is broken?

Submit a controlled test enquiry, confirm the business receives it, and inspect the expected event. Repeat after major form, tag or consent changes.

If your tracking doesn’t match the way enquiries actually arrive, fix the definition and test the journey before trying to optimize the campaign harder.