Check whether contact clicks are distorting your bidding goals

Find the behavior the campaign is rewarded for

An account can report many conversions while staff see few real inquiries if the counted action is a contact-button click or repeated phone tap. Begin by opening the conversion action definitions, not by looking only at the campaign total. A click can help diagnose interest, but it is a different outcome from an accepted form, a connected relevant conversation, or qualified work. The business must choose which stage should guide optimization.

Source checks: About primary and secondary conversion actions · Conversions vs. key events in Google Analytics

Inspect action settings and campaign goals together

Google distinguishes primary and secondary actions, but the goal used by the campaign also matters. Secondary actions included in a custom goal can still be used for bidding. Review account-default goals, campaign-specific selections, and custom goals alongside each action's definition. Record the actual combination. A reassuring secondary label in one screen is not a complete audit of what the campaign optimizes toward.

Create a diagnostic-versus-outcome map

List every action and the evidence it represents: page engagement, contact attempt, delivered request, service qualification, or booking. Mark which business record can verify the latter stages. Keep helpful diagnostic events visible in reports without treating them as interchangeable lead counts. Do not combine a native Ads tag and a GA4 import of the same inquiry into two independent primary successes without reviewing whether they describe one underlying outcome.

Repair the inquiry signal before switching goals

Check successful delivery, duplicate firing, consent behavior, and the normal outcome delay for the action you want to use. A newly named conversion is not reliable simply because it sounds closer to a lead. Test the measurement path with labeled fictional requests and inspect the business store. If qualified outcomes are imported later, verify the import process and definitions. Do not replace one weak signal with another unverified event.

Plan the goal edit as a measurement change

Document current settings, intended new settings, responsible owner, and the reporting date of the edit. Changing optimization inputs can alter campaign behavior, so avoid layering unrelated major targeting and bid changes on the same day unless a definite fault requires repair. Preserve the prior diagnostic counts for context. Make no guarantee that a more meaningful goal will produce a specific cost improvement; it improves the question the campaign is asked to optimize.

Judge the result against actual inquiry quality

After the edit and appropriate outcome delay, compare delivered requests, qualified inquiries, bookings, and spend. Expect the headline conversion count to change when its definition changes. A lower count can be a more accurate report rather than worse performance. Explain the definition break clearly so the owner does not compare the old contact-click total with the new inquiry total as if both measured the same thing.

Worked example: illustrative figures

For example, an old report counts 40 phone taps and 10 delivered forms as 50 primary actions, even though staff confirm only 12 distinct qualified inquiries. The owner retains phone taps for diagnostics and configures the intended inquiry goal after testing it. A subsequent report showing 14 meaningful outcomes cannot be compared with the old 50 as a 72% loss in customers. The definitions differ. The illustration does not assume that every form or phone tap is unique, qualified, or attributable in the same way.

Questions about this task