A Google Ads audit checklist that leads to useful decisions

1. Define the account and the decision

Record the customer account ID, currency, timezone, campaign scope, and reporting period. Confirm which accounts are included and whether any provider read failed or returned truncated results. Keep brand and non-brand activity identifiable when the distinction matters to the business.

Write down the business objective in one sentence. For example: increase qualified demo requests within a monthly spending plan. A form submission, a booked meeting, and a qualified opportunity are different outcomes. An audit cannot judge efficiency accurately if it changes the outcome definition between sections.

2. Check measurement before optimizing traffic

Review the conversion actions used for bidding and reporting in Google Ads. Look for duplicate events, inconsistent values, a recent tracking change, or a low-value event being treated like a sale. For transactions with different values, confirm how those values enter the account. Google explains the role of conversion values.

Account for delayed conversions before judging recent performance. Conversion lag can make recent CPA appear higher and ROAS lower. Choose a mature comparison window appropriate to the buying cycle. Mark recent cohorts as provisional rather than applying an arbitrary universal waiting period.

3. Review comparable campaign performance

Compare equal-length periods with similar weekdays. Show spend, clicks, conversions, and conversion value when available. Calculate aggregate CPA from total spend divided by total conversions; calculate ROAS from comparable value and spend. Do not average campaign ratios or add currencies together.

Separate the size of the change from its explanation. A campaign that accounts for most of the spending increase deserves a closer look. It does not prove that bidding, targeting, or creative caused the change. Identify what additional evidence would distinguish those explanations.

4. Inspect search intent and the destination

Use the search terms report to identify queries with a clear mismatch to the offer. Google describes using search terms to find negative keyword candidates. Separate irrelevant intent from a relevant query that simply has little conversion evidence.

Check the destination page for the offer, location, device experience, and expected conversion path. Confirm that the page promises what the ad promises. Keep a review item for tracking or landing-page issues outside the ad account instead of trying to fix every problem with a bid change. The negative keyword workflow shows how to review exclusions without blocking useful demand.

5. Check budgets and recent strategy changes

Compare actual spend with the client's monthly operating plan, not just yesterday's daily budget setting. Look for campaigns consuming most of the remaining plan, shared budgets, scheduled changes, and unusual changes in delivery. Use the budget pacing guide for a worked example.

Review bidding and structural changes before attributing a short-term result to a creative or keyword problem. Google lists conversion volume, conversion cycles, and strategy changes among learning-period factors. Avoid stacking new changes on top of an experiment before deciding what evidence will count as a result.

6. Reconcile history and policy issues

Check Google's change history for changes made outside Adport. Use Adport's audit history for operations that went through Adport. Neither a chat transcript nor one application's log should be presented as the complete history of every platform action.

Review disapprovals and limited-serving issues in the advertising platform. Note the affected entity, its business impact, and the owner of the fix. A low-spend campaign with an obvious policy issue can need a different response from a campaign that is fully eligible but reaching the wrong people.

7. Produce a small action register

Preview supported advertising changes individually. Review the exact values and policy adjustments before approving the matching operation. After applying, read back the platform state and record the result. This turns the audit into a repeatable operating process rather than another document no one uses.