Triage Google Ads recommendations without auto-applying anything
The recommendations feed read critically: each one is a claim to verify against the account's own data, not an instruction.
IntentGoogle's suggestions sorted into worth-it, ignore, and actively-harmful.
- 2
- Steps
- 3
- Tools
- 2
- Rules
- 1
- Failure modes
When to use it
Use it when
- Monthly.
- When optimisation score pressure arrives from the platform or the client.
Do not use it when
- Never auto-apply: several recommendation types raise spend by design.
What it needs first
Required
Preconditions
- The account is ticked in the dashboard, confirmed via list_accounts.
Inputs
window- Analysis window.Default: last 30 days.
Procedure
In order. Every tool named is one the gateway ships, and links to its reference.
Pull current recommendations with their types.
For each, check the underlying claim against account data before classing it.
Decision rules
The observable condition, what it lets you conclude, and what takes the conclusion back.
When
A recommendation raises budgets or broadens targeting
Conclude
Classify as spend-expansion, not optimisation: evaluate against the efficiency target, not the optimisation score.
When
A recommendation fixes a real gap (disapproved ad, missing sitelink)
Conclude
Worth-it list: cheap, reversible, no spend risk.
Evidence every conclusion must carry
- Every figure carries its account, metric and window.
- Anything below a readable sample size is reported as unjudged, not as zero.
Where the agent stops
- The readout changes nothing. Any change it motivates goes through Safe Writes: preview, human confirmation, then apply.
How it goes wrong quietly
The cases where the analysis is wrong and still looks right. Read them before trusting a number.
- Optimisation score is a platform metric, not a performance metric: reporting it as health misleads clients.
What the answer contains
- 01Readout
- The figures, with windows and accounts named.
- 02Flags
- What deserves a deeper skill or a human decision.