Google Ads Audit: A 20-Point Checklist

Most underperforming Google Ads accounts are not broken in some exotic way. They leak in the same predictable spots: broad keywords with no negatives, conversions that count form-fills instead of qualified leads, and a Performance Max campaign quietly eating budget that the search campaigns earned.

An audit is how you find those leaks before they cost you another quarter. This is the checklist a paid search team actually runs when a new B2B account lands on their desk. Twenty points, grouped the way you should work through them: account health first, then where the money goes, then the parts that decide whether a click becomes a deal.

Set aside two to three hours. Open the account, open a blank doc, and write down what you find next to each point. The goal is not a pretty report. It is a short list of changes ranked by how much money each one moves.

Before you start: pull the right data window

Look at the last 90 days, not the last 7. B2B sales cycles are long and weekly numbers swing too hard to trust. If the account spends heavily, 30 days is fine. Segment by device, by network, and by campaign type early, because the averages hide most of the problems. An account at a 3% conversion rate overall can be running 6% on Search and 0.4% on Display, and the only useful insight is in that split.

Account structure and settings (points 1 to 5)

1. Is the account structured around how you sell, not how Google groups things?

Campaigns should map to your real budgets and priorities: by service line, by margin, by geography you can actually serve. If one campaign mixes high-intent bottom-funnel keywords with broad research terms, you cannot control budget where it matters. A clean structure is boring and it is the thing that makes every other lever work.

2. Are Search and Display separated?

The single most common waste pattern in older accounts: "Search Network with Display select" left on, so a search campaign silently spends 30% of its budget on the Display Network at a fraction of the conversion rate. Split them. Display and Search are different jobs with different economics, and mixing them means you can never read the numbers.

3. Is location targeting set to "Presence," not "Presence or interest"?

The default setting (people in or interested in your locations) lets your ads show to users anywhere who searched about your target city. For a B2B firm serving one region, that is budget poured into the wrong country. Set it to "Presence: People in your targeted locations." Check excluded locations too.

4. Do the campaign settings match the goal?

Ad rotation, ad schedule, network settings, and the bidding strategy should all point the same direction. A lead-gen account on "Maximize clicks" is optimizing for the wrong thing. A common one: campaigns still running 24/7 when the sales team only answers the phone Monday to Friday. You are paying for 3am clicks no one follows up on.

5. Is the account free of obvious orphans?

Paused campaigns with active ad groups, ad groups with one dead keyword, experiments that ended six months ago and never got cleaned up. Clutter is not just untidy, it makes the account hard to read, and an account no one can read is an account no one optimizes.

Where the money goes (points 6 to 11)

This is the part of the audit that pays for itself. You are hunting for spend that produces clicks and no pipeline.

6. Run the Search Terms report. What are you actually paying for?

This is the highest-value 20 minutes in any audit. Sort the search terms report by cost and read the actual queries triggering your ads. In almost every account you will find spend on job seekers ("[your service] jobs"), DIY searchers ("how to do [service] myself"), free hunters ("free," "template," "cheap"), and unrelated terms your match types reeled in. Each one is a negative keyword you should have added months ago.

7. Is there a real negative keyword list?

Most accounts have a thin, reactive list. A B2B account needs a structured one: a shared list for universal junk (jobs, free, DIY, salaries, courses) applied everywhere, plus campaign-level negatives to stop your own campaigns from competing with each other. If the list looks short, that is a finding, not a relief. Our guide to building a negative keyword list walks through the categories that matter most for B2B.

8. What match types dominate your spend?

Broad match without a tight negative list and a smart bidding strategy to guide it is how budgets evaporate. Check how much of your spend sits on broad match and what it actually converted. Broad can work, but only with conversion-based bidding and clean signals feeding it. If broad match is spending freely on a target-CPA strategy that has barely any conversion data to learn from, it is guessing with your money.

9. Is Performance Max stealing credit from Search?

When a Performance Max campaign runs alongside Search, check whether PMax is harvesting your branded and high-intent traffic and getting credit search would have won anyway, often cheaper. Look at whether PMax conversions are mostly brand searches. Exclude your brand terms from PMax if so. For the B2B-specific setup, see when Performance Max actually fits B2B.

10. Are there zombie keywords and ad groups?

Keywords with spend and zero conversions over 90 days deserve a hard look. Some are brand-new and need data. Many are just slow bleeds. Pause or lower bids on the consistent losers. The flip side matters more: find the keywords that convert and are capped by budget or limited by Impression Share, and feed them.

11. What is your wasted spend number?

Add it up. Total cost on search terms you would never have chosen, plus zero-conversion keywords past their learning window, plus the Display leak, plus off-hours spend with no follow-up. Put a dollar figure on it. That single number is usually what gets the account changes approved.

Leak	Where to find it	Typical fix
Display leak on Search campaigns	Segment by network	Split or disable Display Network
Irrelevant search terms	Search Terms report by cost	Add negatives
Off-hours clicks, no follow-up	Ad schedule report	Dayparting to business hours
Brand traffic claimed by PMax	PMax search-term insights	Exclude brand terms from PMax
Zero-conversion keywords	Keyword report, 90 days	Pause or cut bids

Figures and categories above are illustrative; your account will surface its own mix.

Keywords, ads, and Quality Score (points 12 to 15)

12. Do your keywords match real buyer intent?

Group keywords by intent, not just by topic. A search for "[product] pricing" is a buyer; "what is [product]" is a researcher. They should not sit in the same ad group with the same ad and the same bid. Misread intent is why accounts get clicks that never call.

13. Are ad groups tight enough to be relevant?

If an ad group holds 40 loosely related keywords, no single ad can be relevant to all of them, and relevance is what you pay for. Tighter ad groups (a handful of close-variant keywords each) let the ad and landing page speak directly to the query. That lifts click-through rate and Quality Score, and Quality Score lowers what you pay per click.

14. Are you running strong Responsive Search Ads?

Check ad strength, but do not worship it. What matters: at least one strong RSA per ad group, headlines that name the buyer's problem and your differentiator, and pinned positions where your brand or a compliance line needs to stay put. Thin ad copy is a quiet conversion killer. If you only have one weak ad per group, that is a finding.

15. What is Quality Score telling you, component by component?

Do not read Quality Score as a grade. Read its three parts: expected CTR, ad relevance, and landing page experience. Each points at a different fix. Low ad relevance means your ad groups are too broad. Low landing page experience means the click lands somewhere that does not match the promise. We cover the levers in detail in what affects Quality Score and how to improve it.

Conversions, tracking, and the money math (points 16 to 20)

You can fix every keyword and still fail here. If the account is optimizing toward the wrong conversion, every other improvement compounds in the wrong direction.

16. What counts as a conversion, and is it the right thing?

Open Tools, then Conversions. Look hard. Are you counting page views as conversions? Counting every button click? Double-counting a lead across two tags? Many accounts optimize toward "form loaded" or "thank-you page" without checking those translate to real leads. The account will faithfully buy you more of whatever you told it to value. Make sure that is a qualified lead.

17. Is "Include in conversions" set deliberately?

Each conversion action has an "Include in conversions" toggle that decides whether Smart Bidding optimizes toward it. A newsletter signup might be worth tracking but not worth bidding toward. If soft, low-value actions are included, your bidding is chasing volume over quality. Audit that column row by row.

18. Are offline conversions and lead quality flowing back in?

This is the one that separates a clicks account from a revenue account. For B2B, a form-fill is the start, not the sale. If your CRM does not send qualified-lead and closed-deal data back to Google Ads, the algorithm optimizes toward cheap leads, not good ones. Even importing a "qualified lead" status changes which clicks the system chases. No closed-loop data is one of the most expensive gaps an audit finds.

19. Does the bidding strategy have the data to do its job?

Smart Bidding needs conversion volume to learn. A target-CPA strategy on a campaign with three conversions a month is guessing. Check that each automated campaign has enough conversions (a rough floor is around 15 to 30 per month, treat as a guideline) and that the target is realistic against history. If not, the fix may be consolidating campaigns or stepping back to a simpler strategy. Our overview of Smart Bidding and how to train it covers the data thresholds.

20. Can you tie spend to revenue, not just to leads?

The final question, and the one that reframes everything above. Pull cost, leads, qualified leads, deals, and revenue side by side per campaign. The campaign with the lowest cost per lead is often not the one producing deals. If you cannot build that view, the audit's biggest recommendation writes itself: fix the measurement before scaling the spend.

What to do with your findings

A list of twenty observations is not an action plan. Rank them. For each finding, estimate the money it moves and the effort to fix, then sort. In most B2B accounts the order looks roughly like this:

1. Fix conversion tracking and conversion definitions (points 16 to 18). Everything downstream depends on it.
2. Stop the bleeding: negatives, Display leak, off-hours spend, brand cannibalization (points 2, 6, 7, 9).
3. Reallocate budget toward what converts and is capped (point 10).
4. Improve structure, ads, and Quality Score for compounding gains (points 12 to 15).
5. Build the revenue view and offline conversion feedback so the account keeps improving on its own (points 18 to 20).

The structural and copy fixes feel productive, but they rarely move the number as fast as fixing what you measure and where the money leaks. Do those first.

Frequently asked questions

How often should I audit a Google Ads account?

A full 20-point audit quarterly is a sensible rhythm for most B2B accounts. Run a lighter check monthly: search terms, wasted spend, and conversion volume. If you just inherited an account or spend changed sharply, audit now.

How long does a Google Ads audit take?

A focused audit on a single mid-sized account takes two to three hours if the data is accessible. Large accounts with many campaigns and a messy structure can take a full day. The search terms report and conversion setup are where you should spend most of that time, because that is where most of the money hides.

What is the most common problem audits find?

Bad conversion tracking. Either the account counts the wrong actions, double-counts, or never feeds qualified-lead and deal data back from the CRM. When the account is optimizing toward a weak signal, no amount of keyword tuning fixes the core problem.

Can I audit my own account or do I need a specialist?

You can run most of this checklist yourself, and you should at least once, because it teaches you where your account leaks. A specialist tends to spot the subtler issues faster: attribution gaps, bidding-strategy mismatches, and PMax cannibalization. The first pass is the one that always pays off.

Will fixing these things lower my costs?

It usually improves efficiency rather than simply cutting spend. Adding negatives and fixing tracking removes wasted clicks, which lowers your effective cost per qualified lead. The smarter outcome is often spending the same budget on better leads, not spending less.

Does this checklist work for Performance Max campaigns?

Partly. PMax is more of a black box, so several points (tight ad groups, granular keyword control) do not apply the same way. The principles still hold: clean conversion data, brand exclusions, and revenue-based measurement matter more in PMax, not less, precisely because you have fewer manual levers.

The short version

Before you close the account, run down the list:

• Structure maps to how you sell; Search and Display are separate.
• Location set to "Presence"; schedule matches when you can follow up.
• Search terms read and negatives added; wasted spend has a dollar figure.
• Match types and PMax are not quietly burning or cannibalizing budget.
• Ad groups are tight; at least one strong RSA each; Quality Score read by component.
• Conversions count qualified leads, not page views; offline data flows back from the CRM.
• You can tie spend to revenue per campaign, not just to leads.

If you have worked through all twenty and the picture is still murky, or the wasted-spend number turned out bigger than you expected, that is usually the moment to bring in a second set of eyes. Book a 30-minute audit review with Lead The Way: send us access, and we will walk you through the three changes that would move your cost per qualified lead the most. No retainer required to find out.