The Google Ads audit checklist we run before touching a budget
What to check before you blame the agency or raise the budget: structure, matching, negatives, tracking and landing pages, in the order that finds waste.
A Google Ads audit should check conversion tracking first, then the search terms report, then structure, ads, landing pages and settings, in that order. Tracking comes first because Google Ads bids towards whatever you mark as a primary conversion action (Google Ads Help), so a wrong definition of a lead makes every other number wrong too.
The 38 checks below each have a pass condition, so you can mark them pass or fail rather than debate them. In the neglected accounts we open, the waste usually shows up in the first two sections. One is a conversion action counting something that is not an enquiry. The other is searches that have never converted, spending money every day. Raising the budget on an account like that buys more of the same.
The short version
- Check tracking before anything else. Google Ads bids towards primary conversion actions, so a page view counted as a lead teaches the account to buy page views.
- The search terms report shows where the money goes. Read the top 100 rows by cost over 90 days and total what you could not have served.
- Negative keywords do not match close variants. Plurals, misspellings and synonyms need adding separately, according to Google’s own help page.
- The default location setting reaches people outside your area. “Presence or interest” also reaches people who have shown interest in your locations, not only people in them.
- Search partners are switched on by default in a new Search campaign, and auto-applied recommendations can change an account without anyone reading them first.
- The landing page is part of the audit. An ad cannot convert better than the page it sends people to.
What you get
| Format | What it is | Who it is for | Time to use |
|---|---|---|---|
| Checklist | 38 checks in six tables, each with a pass condition and where to look, so the findings can go straight to whoever manages the account. | Anyone spending on Google Ads who cannot explain what the spend returns, or whose cost per lead has climbed without explanation. You need view access to the account, nothing more. | Two to three hours |
What should a Google Ads audit check first?
Conversion tracking. Primary conversion actions are the ones reported in the Conversions column and used for bidding, while secondary actions are for observation only (Google Ads Help). Smart bidding does exactly what that definition asks, so optimising an account with the wrong definition makes it wrong faster.
| # | Check | Pass if | Where to look |
|---|---|---|---|
| 1 | What each conversion action counts | Every action fires on a real enquiry, sale or booked call, never a page view or a form start | Goals, then Conversions, then Summary |
| 2 | Which actions are primary | Only the actions that mean money are primary. Everything else is secondary | Conversion action settings |
| 3 | Duplicate counting | No enquiry is counted twice, for example once by a tag and again by an import | Conversion actions list and their sources |
| 4 | Call conversion length | Call actions only count calls over a minimum length you chose, long enough to exclude wrong numbers | Call conversion action settings |
| 5 | Conversion window | Matches your real sales cycle rather than the 30-day default | Conversion action settings |
| 6 | Conversion values | Values are passed where sales differ in worth | Conversion action settings, Value column |
| 7 | Tag health | At least one primary conversion recorded in the last seven days | Conversions column, last 7 days |
Two of these have rules worth knowing. A call from an ad on your website counts as a conversion only when it lasts longer than a minimum length you set (Google Ads Help). And if you do not customise the click-through conversion window, it is 30 days, with a range of 1 to 90 depending on the source (Google Ads Help). Values matter because value-based bid strategies, such as Maximize conversion value and Target ROAS, optimise for total value rather than a count of conversions (Google Ads Help).
Where is the money actually going?
The search terms report shows the searches people actually typed that triggered your ads, which is different from the keywords you chose. It lets you add any irrelevant one as a negative keyword (Google Ads Help). In our view it is the most useful screen in the platform, and in the accounts we open it is often the one nobody has read for months.
| # | Check | Pass if | Where to look |
|---|---|---|---|
| 8 | Top search terms by cost | You have read the top 100 rows for the last 90 days, sorted by cost | Insights and reports, then Search terms |
| 9 | Waste total | Every term you could not serve is marked, and their cost is added up in currency | Search terms, exported to a sheet |
| 10 | Competitor names | None appear unless you chose to bid on them | Search terms, filtered by name |
| 11 | Job and learning terms | No spend on job, career, salary, training, course, free, DIY or how-to searches | Search terms, filtered by word |
| 12 | Locations | No searches naming places you do not serve | Search terms, filtered by place name |
| 13 | Negative keyword lists | Lists exist and are applied to every campaign that needs them, not just one | Shared library, then Negative keyword lists |
| 14 | Negatives blocking good terms | No negative stops a search you want | Negative keywords, compared against your keyword list |
Row 14 fails more often than people expect. Negative keywords do not match close variants, so plurals, misspellings and synonyms have to be added separately. A negative broad match keyword blocks a search only when the search contains all of its terms, in any order (Google Ads Help).

Photo: Jakub Zerdzicki on Pexels.
Is the account structure wasting budget?
Structure fails quietly. A campaign that mixes a high-value service with cheap enquiries lets the cheap ones take the budget, and the report still looks busy.
| # | Check | Pass if | Where to look |
|---|---|---|---|
| 15 | Campaign split | Campaigns are split by intent and margin, so high-value services do not share a budget with cheap enquiries | Campaigns list and budgets |
| 16 | Match types | Broad match runs only where there is a strong negative list and reliable conversion data | Keywords, Match type column |
| 17 | High click-through, low conversion | No keyword has a strong click-through rate and a poor conversion rate without a known reason | Keywords, sorted by click-through rate |
| 18 | Spend with no conversions | No keyword has meaningful spend and zero conversions over 90 days | Keywords, filtered to 0 conversions |
| 19 | Ad group themes | Every keyword in an ad group fits the same ad | Ad groups and their keywords |
| 20 | Performance Max | You can see which searches it served on and what it spent on brand | Performance Max search terms report |
| 21 | Budget caps | The campaigns with the best cost per conversion are not losing impression share to budget | Campaigns, Search lost IS (budget) column |
For row 20, Performance Max now has its own search terms report, and negative keywords can be applied to its Search and Shopping inventory at account or campaign level (Google Ads Help). The check is whether anyone has opened it.
Do the ads and assets pass?
This section finds fewer savings than the two above, which is why it comes after them. It still decides whether the right person clicks.
| # | Check | Pass if | Where to look |
|---|---|---|---|
| 22 | Ads to compare | Each ad group has at least two responsive search ads, so there is something to compare | Ads, grouped by ad group |
| 23 | Offer in the copy | Headlines name the actual service and offer, not a description of the company | Ad preview |
| 24 | Assets | Sitelinks, callouts and structured snippets are present and specific | Assets |
| 25 | Call asset | The number is answered during every hour the call asset shows | Assets, then the ad schedule |
| 26 | Disapprovals | No disapproved or limited ads sitting unnoticed | Ads, Status column |
| 27 | Ad strength | Ad strength is not being raised at the cost of a message that converts | Ads, Ad strength against conversion rate |
Google’s own best practice for row 22 is at least two responsive search ads per ad group, each with Good or Excellent Ad Strength (Google Ads Help). Row 27 is our caution on top of that: a higher rating is worth having only if the message still says what you sell.
Does the landing page pass?
The page is half the conversion rate and it sits outside the ad account, which is why audits skip it. Click every ad on a phone before you judge it.
| # | Check | Pass if | Where to look |
|---|---|---|---|
| 28 | Answers the search | The page speaks to the specific search, not the company in general | Your own ads, clicked |
| 29 | Speed on a phone | Loads in under three seconds on mobile data. This is our working threshold, not a Google rule | A real phone, off Wi-Fi |
| 30 | Enquiry above the fold | The call or form is visible without scrolling on a phone | A real phone |
| 31 | Form length | The form asks only for what you need to reply | The form |
| 32 | Tappable phone number | Tapping the number starts a call on mobile | A real phone |
| 33 | After submitting | The visitor sees what happens next, not a blank page | Submit a test enquiry |
| 34 | Ad and page match | The page uses the same words as the ad | Ad text beside the page headline |
A mismatch in row 34 tends to show as a good click-through rate and a poor conversion rate, which sends you back to row 17.

Photo: Tima Miroshnichenko on Pexels.
Which settings quietly cost money?
Four settings change who sees your ads without anyone noticing, and two of them are on by default.
| # | Check | Pass if | Where to look |
|---|---|---|---|
| 35 | Location option | Set to “Presence” if you only serve people in your area | Campaign settings, Locations, Location options |
| 36 | Networks | Search partners and Display are on only where you have decided they earn their place | Campaign settings, Networks |
| 37 | Ad schedule | Ads run when someone can answer an enquiry | Campaign settings, Ad schedule |
| 38 | Auto-applied recommendations | Switched off, or reviewed on a schedule | Recommendations, then Auto-apply |
The default location option is “Presence or interest”, which reaches people in, regularly in, or who have shown interest in your targeted locations. “Presence” reaches only people in or regularly in them (Google Ads Help). Search partners are included by default when you create a campaign for the Search Network, and you can opt out in campaign settings (Google Ads Help). Auto-applied recommendations apply regularly once switched on, and can be turned off at any time (Google Ads Help).
Why is the list in this order?
Tracking comes first because every decision after it depends on the data being right. The search terms report comes second because that is where waste is visible in plain numbers rather than inferred.
Cutting waste does two things at once, in our experience. It lowers cost per lead straight away, and it gives the bidding cleaner data to learn from, which tends to lower it again over the following weeks. A common finding in the accounts we open is a conversion action counting something that is not an enquiry, with the platform faithfully buying more of it for months.
The bottom line
An account is only as good as its definition of a lead and the searches it pays for. Check those two before you judge anything else, and write the waste down in currency, because a figure in currency gets acted on and a percentage often does not.
Questions people ask about this
I only have view access. Can I still do this?
Yes, all of it. Every check here is a read. You will need edit access to act on what you find, but the audit itself only requires being able to look, which also makes it a reasonable thing to ask a current agency for.
What if the agency says these settings are deliberate?
Some of them will be, and a good answer explains the reasoning in terms of your cost per qualified lead. If the answer is that it is best practice, or that the platform recommends it, ask what it costs and what it returns. Those are answerable questions.
How much waste is normal?
There is no honest benchmark, because it depends entirely on how long the account has run without a search terms review. We do not publish a percentage because any figure we gave you would be from our accounts, not yours. Add up your own number from section two and judge it against what a customer is worth.
Where do I start if I only have an hour?
Open the conversion actions and read what each one counts, then confirm one has fired in the last seven days. Then open the search terms report for 90 days, sort by cost, and read the top fifty rows. In the neglected accounts we see, those two checks find most of the money. Everything else on the list can wait for a second sitting.
How often should I run this audit?
Fully, once a quarter. The search terms and conversion tracking sections more often, because those are the two that decay fastest. New queries match every week, and a tag can break with a site update on any day. A monthly fifteen-minute pass on those two sections, with the full list quarterly, catches most problems before they cost a whole month of budget.
What if there is no conversion tracking at all?
Stop the audit and fix that before reading any other number. Without conversions, every figure in the account describes clicks rather than customers, and any bidding strategy is optimising blind. Set up at least one action that fires on a confirmed enquiry, such as a thank-you page or a call above a minimum duration. Give it two to four weeks of data before judging the campaigns.
Should I pause a keyword with spend and no conversions?
Check three things first. Does the search term report show good queries under it, or junk? Is the conversion window long enough for your sales cycle? Has the conversion tag been working the whole period? If all three are fine and the spend is meaningful over 90 days, pause it. If the queries look right but the page is weak, fix the page before pausing the keyword.
How do I judge whether Performance Max is worth keeping?
Compare its cost per conversion against your search campaigns over the same 90 days, using the same conversion action. Then check how much of its spend and conversions come from branded searches, which you would have won anyway. If it is cheaper than search only because it is claiming brand traffic, it is not adding much. Add brand exclusions and compare again before deciding.
What is the difference between presence and presence or interest?
Presence reaches people in, or regularly in, your targeted area. Presence or interest, the default, also reaches people who have shown interest in your area, wherever they are. For a service business serving one region, the second setting pays for clicks from people planning a trip, researching a move, or comparing prices from another state. Switch it to presence and watch the location report for a month.
Can I hand this checklist to my current agency?
Yes, and how they respond tells you a lot. A good agency will walk through each section with numbers and explain any setting that looks odd in terms of cost per qualified lead. Reluctance to share the search terms report or the conversion action definitions is the warning sign. Ask for view access to the account so you can check the findings yourself rather than relying on a summary.
What should I do with the waste figure once I have it?
Turn it into negatives first. Every term you marked as unserviceable becomes a negative keyword at the campaign or shared-list level, applied to every campaign. Then decide whether to reinvest the saved budget or reduce spend. Give the account three to four weeks after the change, because smart bidding needs time to relearn from cleaner data before cost per lead settles.
Free review
Want this looked at on your own site?
Send the address. A person reads it and writes back with the two or three things that would move your enquiry number first. Free, and yours to keep whether you hire us or not.