Google Ads account audit and Quality Score guide: what to check in six areas

Quality Score is a 1 to 10 rating Google gives each keyword by combining expected click-through rate, ad relevance and landing page experience. As the score falls you must bid higher to hold the same position, so a low score means a more expensive click.

The goal of an audit is not to lift scores one by one but to find the structural causes that pull them down. In most accounts the same three or four causes repeat.

Structure: are there unrelated keywords inside ad groups, do dozens of keywords in one group point at a single ad, are brand and generic terms separated? A messy structure is the first cause of low ad relevance.

Keywords: list the ones scoring below 5 with high spend. Google tells you per component which of them suffer from low expected CTR and which from a weak landing page. In Tecrube, “list keywords with Quality Score below 5 with their components and spend” is one question.

Search terms: terms that clicked without converting in the last 30 days and off-intent patterns not yet negated. Every term found here leaks both budget and Quality Score, because irrelevant impressions depress expected CTR.

Ad copy: does every ad group have at least two responsive search ads, is ad strength “good” or “excellent”, do the headlines contain the keywords, are any ads disapproved or limited? Missing extensions count under this heading too.

Conversion tracking: which conversion actions are defined, did each receive data in the last 30 days, is the primary and secondary split correct, is the same action counted twice? Broken measurement blinds Smart Bidding; this is the most critical area of the audit.

Budget: which campaigns are limited by budget, are there campaigns that never spend, does the allocation match the campaigns that convert? Moving budget from a capped winner to a campaign that does not spend is the fastest return an audit delivers.

Walking through six areas by hand takes more than half a day. With the Tecrube connector you tell Claude or ChatGPT “audit my account across six areas and prioritise the findings”; the result arrives as a ranked list of findings rather than a pile of data. An audit is a read, spends no credits and needs no approval.

Fixing the findings is a write: adding negatives, moving budget, creating a new RSA. Each follows preview → approve → apply; single-step budget increases above fifty percent are rejected and every change is written to the audit log.

A full audit once a month and a search term and budget check once a week are enough. The 75 credits on the free account cover the first audit and a few fixes; Pro includes 5,000 credits a month and the account hygiene audit is part of the plan. On the Agency plan the audit runs per client account with the same command.

Comparing audit results with the previous month shows in numbers how average Quality Score and wasted spend have moved, which is the most concrete progress report you can give management or a client.

An audit usually produces dozens of findings, and trying to fix them all on the same day is risky and unnecessary. The suggested order: measurement first, because broken conversion data misleads every other decision. Then budget leaks, meaning search terms and campaigns that spend without converting. Structure and ad copy come last, because their effect shows over weeks.

Ask the assistant to table the findings with “estimated impact” and “effort” columns. High-impact, easy items are week one's work, such as adding ten off-intent terms to the shared negative list. High-impact but hard items, such as rebuilding the account structure, need a separate plan.

Waiting two weeks after each fix and rerunning the same audit shows which change worked. Because an audit is a read, these reruns spend no credits.

Asking the assistant for a side-by-side table of two audits is the easiest way to see progress.