The Google Ads Audit Checklist for 2026: Seven Checks That Stop Wasted Spend

Most Google Ads accounts aren't losing money because the strategy is wrong. They're losing it because a quarter of the budget goes to people who were never going to buy, and nobody has looked.

Google won't point it out. Its reports are built to show you what worked, not what leaked.

I do a Google Ads audit for a living, across retail, healthcare, legal and e-commerce in the Gulf and the US, and the same leaks show up almost every time. You don't need a paid tool to find them. They're sitting in reports you already have, and an afternoon is usually enough. This is the order I work in, with what each check turned up on real accounts this year.

1. Read the search terms report like a buyer would

Go to Insights and reports, then Search terms. Last 30 days, sorted by cost. Now read the top 100 rows, one at a time. Don't skim. For each one, ask whether this person wanted what you sell or just something that shares a word with it.

Navigational searches are the first thing I look for. In September I audited a law firm whose most expensive search term was the name of a government reporting website, pulled in by one of their phrase keywords. One row, and the whole problem was visible. Roughly a third of their 98 keywords were catching people looking for a public website or a case status page. Not one of them was going to ring a lawyer. We paused 47 keywords and added 142 negative keywords in a single sitting, and the account dropped to 65 keywords, each naming a lawyer or a live legal problem.

Then definitions. "What is", "meaning of", "how to report". Someone typing "what is a family lawyer" is at least two steps from hiring one, and on phrase match you pay for that curiosity.

Competitor names are the sneaky ones, because they look commercial. They are commercial. The person has just already picked somebody else. On a pet retail account in Saudi Arabia, three phrase keywords were taking 87 percent of the non-brand search budget, 1,045 riyals of 1,206 for the period, on searches naming rival stores. Zero purchases. On paper those keywords looked fine. The search terms said otherwise.

On most accounts this check alone recovers 15 to 35 percent of search spend. That money moves to queries that actually sell, so your cost per lead drops without touching the budget.

2. Match types, and my view on broad

I'm not anti broad. I'm anti broad on accounts that haven't earned it.

Without conversion history and a maintained negative list, broad match is basically a donation to Google. It'll stretch your keyword to anything it thinks is related, and its idea of related is generous.

My rule: broad only on campaigns with 30+ conversions a month, on Smart Bidding, with negatives reviewed in the last fortnight. Everything else runs exact and phrase. That law firm now has no broad keywords at all, 24 exact and 41 phrase, and it's the cleanest search terms report I've seen this year.

Tighter matching usually shows up first as a better conversion rate. Same spend, more leads. That's the cheapest Google Ads ROI improvement there is.

3. Performance Max placements (the report nobody opens)

Honestly, this one annoys me. Most advertisers running Performance Max have no idea where their ads appear, and the answer is one click away.

Open the campaign, go to Insights and reports, and find the placements view. It lists every site, app and YouTube channel that served your ads. Set 90 days, sort by cost, read the domains.

On a medical clinic account in Florida, that list ran to 625 websites and apps since 1 July. Word unscrambler sites. Gematria calculators. People search directories, low-income housing listings, job help forums. Between them they'd spent 3,407 dollars and produced zero sales.

It got worse. They had produced leads, just the wrong ones: Spanish-language enquiries from in-app browser traffic on Latin American sites that the front desk couldn't serve. And the algorithm was learning from them, because to a bidding model a lead is a lead.

Placement exclusions on that account? None. I still see this every month; it's the default state of most Performance Max placements I review.

Build an account-level exclusion list. Start with everything that's obviously not your customer, then add app categories that never convert for you. While you're there, check whether Search Partners and Display expansion are on, and switch them off anywhere the job is catching high-intent search. I typically see 10 to 25 percent of the PMax budget come back, and for a clinic the cleaner lead mix matters as much as the money. Staff stop chasing enquiries that were never going to book.

4. Find out what your bids are actually chasing

If you only do one check, do this one. It costs nothing.

Goals, then Conversions, then Summary. Look at which actions are marked Primary. Those are what Smart Bidding optimises toward. Secondary actions are only watched.

What I find again and again is the money action, the purchase or the qualified lead, sitting as secondary while something soft like a phone number click or an app open is primary. The algorithm does exactly as it's told. It finds people who click phone numbers and doesn't care if they ever buy.

That pet retailer had every app purchase set to secondary and a batch of engagement actions set to primary. The campaigns were bidding for the wrong thing and everyone was blaming the creative. Two minutes to fix. It's the most common root cause I find behind "our ROAS collapsed and we don't know why".

A warning, though, because I nearly got this wrong myself. Campaigns can override the account default, so open each campaign's own conversion goals too. On one audit the account view made it look like bidding was chasing junk. The campaign view showed three of four campaigns were already purchase-only. Check the setting that controls the behaviour, not the one that's easiest to find.

Spend doesn't change here. Return on ad spend does, sometimes dramatically, because the same budget is now pointed at buyers.

5. Location settings

Campaign settings, Locations, Location options. If it says "Presence or interest", you're paying for anyone in the world Google thinks is interested in your city. It's the default. For a local business it's usually wrong; switch to Presence.

Then pull the locations report by user location, not target location. Serving one city and seeing 30 percent of clicks from elsewhere? Those clicks were never going to become customers.

6. Device and hour of day

Segment 30 days by device and by hour. You're looking for real spend with nothing to show for it. Mobile at 2 a.m. for a B2B service is a classic.

Don't cut on a hunch. Wait for 30 days of evidence, and if the segment converts now and then, lower the bid instead of excluding it. The sums here are smaller than checks 1 to 4, but it's clean money.

7. Test one conversion yourself

Submit a real test lead, or place a real test order, and confirm it lands once, with the right value and currency, in Google Ads and in your analytics.

I've seen a Saudi retailer's app conversions mapped through US dollars, which made ROAS look six times better than it was. I've seen accounts count every purchase twice because two events were both primary. Everything above depends on that conversions column. If it's wrong, you're scaling the wrong campaigns with total confidence.

The afternoon, in order

1. Search terms, top 100 by cost, read by hand. Pause, add negatives.

2. Match types. Justify every broad keyword or move it to phrase.

3. PMax placements, 90 days. Build the exclusion list.

4. Primary conversions, at account and campaign level.

5. Presence only for anything local.

6. Device and hour, on 30 days of data.

7. One test conversion, traced end to end.

Most Google Ads tips are about the newest feature. Most real Google Ads optimization is this list, repeated. I'd run a quick PPC audit like this every two weeks.

After the first pass, leave the account alone for two weeks so Smart Bidding can learn from the cleaner signal. Book the next audit for the day that fortnight ends.

FAQ

How long does a Google Ads audit take?

The seven checks above take an afternoon on a typical account. Reading the top 100 search terms by hand is the slowest part, and the most valuable.

How often should I audit a Google Ads account?

Every two weeks for the search terms and placements checks. After a big cleanup, leave the account alone for two weeks so Smart Bidding can learn from the cleaner signal, then audit again.

Do I need a paid tool to audit Google Ads?

No. Every check here uses reports already inside Google Ads: search terms, placements, conversion goals, location, device and hour of day.

What is the single most important check?

Which conversion actions are set to Primary, at account level and inside each campaign. If bidding is chasing the wrong action, nothing else in the account can fix the results.

How much wasted spend does an audit usually find?

On most accounts the search terms check alone recovers 15 to 35 percent of search spend, and cleaning Performance Max placements typically brings back 10 to 25 percent of PMax budget. Your account may land outside those ranges.

Where to go from here

If you'd rather have a second pair of eyes on the account, Big AL is a performance marketing agency in Dubai and a Google Partner since 2021. We run this checklist as part of every Google Ads management engagement in Dubai and every account we take on as a PPC agency in Dubai. The goal is the same each time: the same budget pointed at buyers, a lower cost per lead and more revenue from the spend you already have.