In Google Ads, primary conversion actions can guide bidding when their standard goal is used by the campaign. Secondary actions normally provide observation rather than bidding signals. The important exception is a custom goal: actions included in it are used for bidding regardless of their primary or secondary label. Google's conversion action documentation.
That distinction matters because measuring an action and asking a campaign to optimize for it are different decisions. A click on a contact button is not the same outcome as a completed inquiry.
Map the business outcome before touching settings
Write a short event map showing what the visitor does, what your website records, and what the ad account receives. Avoid labels such as “lead conversion” unless you can describe the exact trigger.
A hypothetical service business might track these actions:
| Action | What it tells you | Question before using it for optimization |
|---|---|---|
| Contact button click | Someone opened the contact path | Did they actually submit or call? |
| Successful inquiry | A request reached your system | Is it valid and deduplicated? |
| Qualified lead | The team confirmed business fit | Is qualification consistent and timely? |
| Won customer | A sale was recorded | Are values and outcome dates reliable? |
This is an audit framework, not a universal instruction to choose the deepest event. A downstream event that is rarely or unreliably reported may not support the intended campaign strategy.
Inspect the campaign's actual goals
Do not stop after reading the action label. Inspect which goals the campaign uses, including campaign-specific settings and custom goals. Google's documentation distinguishes the Conversions reporting column from All conversions, so a difference between those totals can be intentional. Primary and secondary reporting behavior.
Create a small inventory with the action name, trigger, data source, goal, counting approach, and owner. Add the last time someone verified it. A precise inventory is more useful than a screenshot showing that tracking is “active.”
Look for duplicate business events from separate implementations. For example, a native website event and an analytics-imported event might both describe the same submitted form. Investigate before deleting either; one may serve reporting needs even if it should not be counted toward the same optimization objective.
Test the event, not just the button
Use a controlled test inquiry and confirm the sequence from the form to your backend and measurement tools. Where supported, use diagnostic or test modes so QA does not contaminate live business reporting.
Try a failed validation, a successful submission, and a page refresh. A rejected form should not produce the same outcome as an accepted one. A refresh should not silently create another CRM inquiry. Advertising attribution will not necessarily appear for a direct test visit; verify delivery and business behavior separately from attribution.
Record what the test proves and what it does not. Seeing a tag fire does not prove that the lead was saved, assigned, or qualified.
Change deliberately and keep a comparison record
If you correct a goal configuration, record the previous setting, the new setting, the reason, and the change time. Expect the meaning of reported conversions to change if the underlying event set changes.
For instance, moving from button clicks to accepted inquiries can reduce the conversion count without reducing actual customer demand. Explain that break in the reporting series rather than treating it as an automatic campaign failure.
How Omesta helps with checking the goal behind a result
Ask about the campaign and the conversion signal behind its reported performance before treating a low cost per result as success. Connected reporting gives the review a shared starting point. You still inspect and test the actual Google Ads goal configuration; a connected account is not proof that the correct event is primary.
Explore the AI ad-management workflow for current capabilities and the pricing page for plan limits.
Frequently asked questions
Are secondary conversions always ignored by bidding?
No. The custom-goal exception matters. Inspect goal membership as well as the action's label.
Should every valuable action be primary?
Not necessarily. Decide what the campaign should optimize toward, and avoid treating several steps in one journey as interchangeable successes.
Can I compare CPA before and after changing goals?
Only with a clear explanation of the measurement change. If the denominator represents a different action, the two CPAs are not directly equivalent.
For the surrounding terminology, see our CPC, CPA, and CPM guide. Explore Omesta for your advertising workflow.
