How to audit a Google Ads account you just inherited

Almost every account that lands on my desk comes with the same sentence: we're spending and leads aren't coming in, and nobody can tell us why.

From there you can work two ways. One is to start moving bids on day one, because that's what people expect from you. The other is to spend the first few hours understanding what the account is actually measuring before you change anything.

I work the second way. It isn't caution. In most inherited accounts, the problem isn't the bids.

1. Start with whatever the account calls a conversion

Before you open a single campaign, go to Goals and look at the conversion actions.

What you're looking for:

Duplicated actions. It's common to find the same form submission counted twice, once by a native tag and once by an Analytics import. The account shows double the conversions it has, and the bidding algorithm is learning from inflated data.

Primary actions that shouldn't be primary. Primary actions are the ones that feed the conversions column and drive automated bidding. If a visit to the contact page is in there, or a phone click nobody ever verified, the account is optimizing toward an event that isn't worth money.

Conversion windows nobody reviewed. The default is thirty days from the click. On a three-month sales cycle, that cuts off real conversions. On an impulse-purchase ecommerce account, it stretches them too far.

Actions with no conversions in months. If something hasn't recorded anything since March, it's either broken or the page is gone.

When I find an account with broken measurement, I say so on day one and I don't touch bids until it's fixed. Optimizing on bad data means going faster in the wrong direction.

2. The bidding strategy, and whether the account can support it

Look at the bidding strategy on each campaign and, right next to it, how many conversions that campaign records per month. Those two numbers have to be read together.

Maximize conversions needs data to work. At three or four conversions a month the system has nothing to learn from: it spreads the budget close to blind, and you pay for the experiment. As a rule of thumb, below roughly thirty conversions in thirty days, automated bidding is running on fumes.

On new accounts, or accounts with thin volume, manual CPC is still the sensible call. It lets you see which keywords and which ads bring in something real, and meanwhile you build the history that automated bidding needs to do its job. Once that history exists, you switch.

The case I run into most is always the same: a small account set to maximize conversions from day one because that's what the interface recommended, and six months later nobody can explain why cost per lead swings with no pattern.

3. Search terms come before keywords

Keywords tell you what someone intended to buy. Search terms tell you what is actually being bought.

Pull the last ninety days, sort by cost, and keep scrolling until you run out of patience. In an account nobody has maintained, you'll find a share of the spend going to searches with no connection to the business.

This has gotten worse since broad match and automated bidding started working together. The system explores, and exploring costs money. A properly built negative list is one of the few things that gives budget back the same day.

4. Where the ads are actually showing

Two reports almost nobody opens in an inherited account, and where the wasted money usually hides.

Devices. In the device breakdown, tablet almost never performs. The normal picture is spend and clicks with a token conversion or none at all. If nobody has looked at this in months, that's budget sitting unused.

There's a side effect people miss. A large share of tablet Display inventory is games and children's apps. A click there isn't buying intent. It's a finger landing in the wrong place. If the account runs Display and nobody has touched the device settings, there's a good chance you're paying for that every month.

Placements. Pull the placements report and sort by cost. You'll see domains you don't recognize, recycled-content aggregators, and pages built for the sole purpose of selling ad space. That share of the budget leaves every month until somebody sits down and builds the exclusion list.

Check as well whether the Search campaigns have the Display Network and search partners switched on. Both are on by default and they stay on for years.

5. Check whether recommendations are applying themselves

Under Recommendations, the auto-apply tab. This is where I've found the most surprises.

If it's on, Google may be adding keywords, changing bid strategies, or creating ads with nobody approving any of it. Plenty of accounts have it switched on from the time someone passed through two years ago.

It isn't that everything it suggests is wrong. It's that if you don't know which changes were applied, you can't explain why the account behaves differently this month.

6. The structure tells you who built it

I'm not looking for perfect structure. I'm looking for signs of how fast it was thrown together.

Ad groups with thirty keywords inside. Duplicated campaigns named "copy of". Budgets split evenly across campaigns that perform nothing alike.

All of that gets fixed. What matters in the audit is counting it, so you can decide what to fix first.

7. How many creatives are really running

Count the active ads per ad group. Then look at how long they've been up.

An ad that hasn't been touched in eight months against the same audience isn't stable. It's exhausted. In small accounts this shows up less because reach is low. In accounts with volume, fatigue explains performance drops that get blamed on bidding.

8. The landing page is part of the account

This is where I get the most pushback, because the landing page usually belongs to another team.

It doesn't matter whose it is. If the ad promises one thing and the page says another, the click is paid for and lost. If the form asks for nine fields to download a PDF, no bid adjustment is going to fix that.

I check three things: whether the ad's message appears on the page, how many fields the form asks for, and how long it takes to load on mobile.

9. What not to touch in the first week

This matters as much as everything above.

Don't change the bid strategy on day one. Every change restarts the learning period and leaves you two weeks with nothing readable.

Don't pause campaigns in bulk. Understand what each one contributes first, even when it looks like nothing.

And don't make fifteen changes at once, because afterwards you won't know which one worked.

What I hand over

A short document with three parts: what's broken in the measurement, where the budget is going, and what I'd do over the next four weeks in order of impact.

No screenshot that doesn't explain something. No generic recommendations that would fit any account.

If you want this for your account, I run fixed-scope audits for $590, with a recorded walkthrough and a written document. Here's how it works.