Google Ad Grants not getting impressions? Start here.

First, separate “no delivery” from “poor results”

Zero impressions, low click-through rate and few donations are different problems. Choose the reporting period and verify that you are looking at enabled campaigns in the correct grant account. A report filtered to a paused campaign can make an active account look broken.

If there are useful clicks already, start investigating query relevance and the landing-page journey. Repeatedly changing bidding to solve an unrelated form problem usually makes the evidence harder to interpret.

Check the account and eligibility

Compare the ten-digit Customer ID with the account supplied after Ad Grants activation. An organisation may also have a paid account under the same email. Do not add a card to an unfamiliar account to try to unlock a grant.

• Check activation and outstanding account notices.
• Read campaign, ad and keyword status explanations.
• Check campaign dates, schedules and website availability.
• Record the exact reason for a disapproval before requesting a review.

Inspect the goals and bidding

A conversion strategy needs a dependable definition of a useful action. Open the campaign’s selected goals, then test that the relevant action actually records. A linked GA4 property does not by itself establish working conversion actions.

My usual starting point is Maximize Conversions with meaningful tracking and no unnecessarily tight CPA target. That is a working recommendation, not a promise that changing one setting will produce impressions. If goals are broken, fix them first.

Look for restrictions that compound

A narrow service area, highly competitive donation-only keywords, restrictive audience targeting and broad negative lists can combine to leave very little eligible demand. Review them together.

• Keep geography connected to where the nonprofit can genuinely serve or inform people.
• Check whether an audience uses Targeting when Observation was intended.
• Inspect inherited exclusions and negative keywords.
• Check whether useful pages have accidentally been excluded from automated coverage.

Expand demand you can answer well

A community garden might struggle with “donate to charity” yet have useful pages about volunteering, school visits and local growing advice. Each theme needs an honest ad and a page that answers the search. Adding generic news terms with no mission connection is not a sensible expansion.

Broad match can discover additional wording. AI Max can extend Search matching. Performance Max can explore demand using goals, assets and site coverage. Test an appropriate lever, record the date and inspect what it creates. If every lever changes at once, you will struggle to identify the useful one.

Allow learning, then make a decision

After fixing a real blocker, allow the campaign to pass review and accumulate activity. Use campaign reports and status notices as evidence; a single Ad Preview result is not a reliable delivery audit.

If delivery remains absent, prepare the Customer ID, campaign IDs, date range, status messages, goals, targeting and changes already tried for the official support route. Rebuilding the same campaign repeatedly rarely adds diagnostic value.

Should you try a very high target CPA?

Only as an advanced, monitored experiment after ordinary constraints have been excluded. A permissive target does not change programme rules or guarantee better traffic. Define the quality threshold and rollback condition before testing. More grant usage alone is not a successful outcome.