Recommendations backed by rules and evidence
Each audit rule evaluates a defined condition against account data. When a rule triggers, the resulting finding carries a severity, a category, the measured value, the threshold it crossed, structured evidence, a data-confidence label, and the rule that produced it. No recommendation appears without that trail. These rules are the diagnostic engine behind your Google Ads Health Score — every category score on the dashboard is a sum of findings from this registry.
Search
7Search terms with cost and no conversions, keyword-level waste.
Conversion tracking
5Duplicate or missing primary conversion actions, offline import gaps.
Budget allocation
4Budget concentrated in underperforming campaigns, spend with zero conversions.
Performance Max
3Performance Max asset groups missing the recommended minimum assets.
Account hygiene
2Campaigns and ad groups missing basic setup — paused-with-budget, empty groups.
Ads & assets
2Ad groups running a single ad variant with nothing to test against.
A sample from the registry
Real rule identifiers, so a finding on your dashboard can always be traced back to the rule that produced it.
No active conversion action
The account spends budget but no conversion action is active and recording conversions — optimization is running blind.
Too many conversion actions marked as primary
Micro-conversions (add-to-cart, page views) are mixed in with real goals as "primary," diluting the signal Smart Bidding optimizes toward.
Costly search terms with zero conversions
Search terms with spend above a CPA-derived threshold and no recorded conversions — a direct, measurable source of waste.
Campaign spends budget with zero conversions
A campaign has spent a large multiple of target CPA (or a meaningful share of account cost) without a single conversion.
Incomplete asset set in a Performance Max group
An asset group is missing Google's recommended minimum headlines, descriptions, images or logo.
Campaign enabled but effectively inactive
A campaign is marked enabled but has not served in a way that would let it spend — an easy-to-miss configuration gap.
Insufficient data is a distinct, named state
A recommendation based on a handful of clicks or conversions is worse than no recommendation. Every rule has a minimum data volume before it will produce a finding. Below that volume, the rule is skipped and reported separately — it is never silently treated as “no problem found.”
Enough volume for a full-confidence recommendation.
The pattern is visible but the sample is small — the finding is flagged as lower confidence, not withheld.
Not enough data to judge. The rule is skipped and listed as “not checked,” never scored as a pass.
A category with no applicable campaigns (for example, no Performance Max campaigns on the account) is also reported as “not applicable,” distinct from “checked and clean.” Skipped and not-applicable rules never lower the account score.