How to Do a Google Ads Audit: Step-by-Step Checklist

To do a Google Ads audit, work through the account in dependency order: verify conversion tracking first, then campaign structure, keywords and match types, negative keyword coverage, the search terms report sorted by cost, Quality Score on top spenders, bid strategy health, ad and landing page performance, and finally budget allocation against results. Doing it in that order matters, because if tracking is wrong every judgment you make afterward is built on bad numbers. A full audit takes two to four hours on a mid-size account.

Why audit order beats audit length

Most published audit checklists run 50 to 100 items, which is useful as a reference and terrible as a plan. If you start at the top and work down, you will spend your first hour on ad copy variations and never reach the conversion action that has been double-counting sales since March. Everything in a Google Ads account is downstream of measurement, and measurement is downstream of what you told Google to count.

So the sequence below is deliberate. Each step is only trustworthy if the step above it passed. Accounts that have never been audited properly commonly waste 20% to 40% of their budget, and in practice most of that waste traces back to two or three findings near the top of this list, not to the long tail of small optimizations at the bottom.

Step 1: verify conversion tracking before touching anything else

Open Tools, then Conversions, and look at what is marked as a primary conversion action. It should be a real business outcome: a purchase, a submitted quote request, a booked call. If page views, button clicks, time on site, or newsletter signups are sitting in the primary column, your Smart Bidding has been optimizing toward things that do not pay you.

Check three specific failure modes. First, duplicates: the same sale counted by both a Google tag and a Google Analytics import, which inflates conversions and quietly halves your reported CPA. Second, the counting setting: "every" versus "one" conversion, which matters enormously for lead gen where one person submits a form three times. Third, whether the tag still fires at all. Conversion tracking breaks silently after a site deploy, and nobody notices until a monthly report looks strange. Nothing throws an error; the account just stops telling you the truth.

If tracking is broken, stop the audit. Fix it, let two weeks of clean data accumulate, then come back. Auditing performance on bad data produces confident, wrong conclusions.

Step 2: read the search terms report sorted by cost

This is where the money actually goes. Pull the last 90 days of search terms, sort by cost descending, and read the top 100 rows. You are looking for two things at once: queries you should never have paid for, and queries that are converting well but are not yet keywords of their own.

The waste patterns repeat across almost every account. Job seekers ("plumber jobs near me"), students and researchers ("what does a media buyer do"), free hunters ("free" plus anything), competitors' brand names when you are not deliberately bidding on them, and queries that share one word with your product and nothing else. Add these as negatives at the right level: campaign level for anything specific, account level for the universal ones like "jobs", "salary", "free", "career", and "how to become".

On the other side, any search term with conversions and a sensible cost per conversion should be promoted into its own exact match keyword so you can bid on it deliberately rather than catching it by accident. This one step, repeated weekly, is worth more than most of the rest of the audit combined. Our guide to lowering cost per click covers the bidding side of the same problem.

Step 3: check structure, match types, and campaign overlap

Look at how campaigns are organized and whether the organization still reflects how the business makes money. The common findings are a single campaign holding every service the company offers, ad groups with 40 loosely related keywords sharing one ad, and broad match keywords running under a Smart Bidding strategy that does not have enough conversion data to steer them.

Then check for cannibalization. Use Auction Insights and look for your own campaigns competing in the same auctions, which usually shows up between a Search campaign and Performance Max. PMax without brand exclusions and account-level negatives will happily absorb your branded search, report those conversions as its own, and charge you more for traffic that would have converted anyway. Fencing PMax properly is one of the highest-value fixes in a 2026 audit. Our breakdown of how to structure a Google Ads account goes deeper on the layout itself.

Step 4: run the numbers against 2026 benchmarks

Once the account is measured correctly and the obvious waste is blocked, compare the numbers to what a healthy account looks like. These are the 2026 reference points worth checking against.

The impression share split is the one people skip and shouldn't. Impression share lost to budget means you are leaving profitable volume on the table and should spend more. Impression share lost to rank means your ads are not good enough or relevant enough to win the auction, and spending more will not fix it. They look similar on a dashboard and call for opposite responses.

Step 5: match budget to results, not to history

For each campaign, calculate its share of total spend and its share of total conversions or revenue. Put the two columns side by side. You are looking for the campaign taking 40% of the budget and returning 12% of the conversions, and for the campaign returning 30% of conversions while capped by a budget it hits every afternoon.

This is almost always where the largest single improvement in an audit lives, and it is almost always the result of nobody revisiting numbers set months earlier. Budgets ossify. A campaign gets $2,000 a month because that is what it got last quarter, not because anyone checked what it earned. Wasted ad spend behaves exactly like untracked cloud and SaaS spend: it accumulates quietly, in small monthly increments nobody owns, until someone finally sits down and reads the line items.

Work out what each campaign should be allowed to spend from your target cost per acquisition and your customer value rather than from precedent. The advertising budget calculator works backward from a revenue goal to a defensible monthly number, and the CPA calculator gives you the ceiling you cannot exceed and stay profitable.

Step 6: ads, assets, and landing pages

Last, because it only matters once the account is buying the right traffic. Check that every ad group has at least two responsive search ads, that the asset level ratings are not all "Low", and that pinned headlines are not choking the combinations Google can test. Look for ad groups still running a single ad written two years ago.

Then click your own ads. Load the landing page on a phone. Confirm the page matches the promise in the ad, that the form works, that the page loads in under three seconds, and that nothing 404s. A surprising share of "the ads stopped working" cases are a landing page that quietly broke, and no amount of bid tuning compensates for a checkout that fails on mobile.

How often should you repeat this?

Run the full sequence quarterly. Add a lighter monthly review once you spend around $5,000 a month or more. Review search terms and negatives weekly on any active account, because that is the fastest-accumulating source of waste. Audit immediately after any major change: a new bid strategy, a site migration, a tracking update, or a new campaign type going live.

The honest problem with the quarterly rhythm is that it is a compromise with human availability, not a match for how fast an ad account changes. Competitors enter the auction weekly. Tags break on a Tuesday deploy. Budget caps bind on the days demand spikes. That gap between how often accounts break and how often people check them is exactly what our Google Ads audit approach is built to close: the same checks run continuously as software, with the fixes applied rather than written up. If you would rather hand the whole account over, our AI Google Ads management page covers how that works.

Frequently asked questions

How long does a Google Ads audit take?

Two to four hours for a mid-size account with a handful of campaigns, if you know the platform. Large accounts with shopping feeds, multi-location targeting, and thousands of keywords take a full day or more. A quick 15-minute health check covering search terms by cost, network segmentation, hourly performance, auction insights, and bid strategy status will surface the biggest problems in most accounts.

What is the most common finding in a Google Ads audit?

Broken or misconfigured conversion tracking, followed closely by missing negative keywords. Between them they account for most of the waste in a typical never-audited account. Tracking problems are more damaging because they corrupt every other decision, while negative keyword gaps are more expensive per day because the meter runs continuously.

Can you audit a Google Ads account without access to it?

Not properly. Free graders and competitor tools can estimate spend and show which ads are running, but they cannot see conversion actions, Quality Score, search terms, or budget pacing, which is where the real findings are. A genuine audit needs at least read-only access to the account.

Is a free Google Ads audit worth doing?

As a fast sanity check, yes. As a substitute for a real audit, no. Free audits are almost always automated script-based reviews built to start a sales conversation, and they surface surface-level findings by design. They will not catch a conversion action counting the wrong thing or a PMax campaign cannibalizing your branded search.