Google Ads · Audit
Back to Google Ads guides
Quick answer
A business owner can audit a Google Ads account in 90 minutes without agency access by running six checks in sequence: account access, conversion tracking integrity, the search terms report, match type distribution, campaign structure, and Performance Max asset groups. The goal is not optimization. The goal is to surface the three to five structural problems that are quietly costing the account 20 to 40 percent of its monthly budget.
Key takeaways
What this article covers
A business owner reviewing a Google Ads account for the first time almost always starts in the wrong place. They open the account, look at the top-line numbers, see CTR and conversion rate, and try to decide if the account is doing well. That path yields nothing. The numbers on the dashboard are the output of decisions made elsewhere in the account. The audit starts at the decisions, not the output. This guide is the sequence I use after twenty years of running paid advertising across US, European, and Asian markets. It works on any account, whether you set it up, inherited it, or pay an agency to manage it. For the full pillar, see the Google Ads guides collection.
Across hundreds of audits, the same five or six problems show up in roughly the same order. Not because the platform is defective. Because most accounts were built by someone optimizing for the wrong objective, or set up in a hurry by a contractor who treated the build as a deliverable rather than an asset. Once an account is running, nobody goes back to the architecture. Bids get tweaked, budgets get raised, new ad copy gets written. The foundation never gets revisited. Six months in, the account is layered with decisions that each made sense in isolation and collectively produce a structure that drains money.
The patterns are consistent enough that you can predict what you will find before you log in:
Each of these is structural. None of them shows up on the main dashboard. All of them compound.
Every audit starts here. Not because conversion tracking is the most interesting finding. Because an account optimizing toward a broken conversion event gets worse every day the algorithm runs. Smart Bidding reads the conversion signal and adjusts. If the signal is wrong, the bids are wrong, and the longer the campaign runs, the further off it gets. Fixing tracking on a six-month-old campaign is not the same as starting a new campaign with correct tracking, because the algorithm has already trained on bad data.
Open Tools, then Conversions. For every primary conversion action, verify:
If the conversion action is the Shopify Purchase event, verify it against the actual Shopify order count for the same date range. A 10 percent variance is normal. A 40 percent variance means the tag is firing in the wrong place or double-counting.
The Search terms report is the single most valuable diagnostic in Google Ads and the report most operators have never opened. It shows the actual queries that triggered your ads, not the keywords you thought you were bidding on. In a healthy account, you will recognize 80 percent of the top-cost queries as real buyers. In a broken account, you will recognize less than half.
The procedure is mechanical. Go to Insights, then Search terms. Change the date range to the last 90 days. Sort by cost, descending. Read the top 30 queries out loud. For each query, decide: is this a person my business wants to pay to reach? If not, flag it. Common patterns that indicate a problem:
The top 30 queries typically account for 50 to 70 percent of total spend. A single afternoon adding negatives often recovers 15 to 25 percent of monthly budget without touching anything else.
Match type is the single decision with the largest budget impact and the one most accounts get wrong by default. Google's interface pushes broad match as the recommended setting because broad match combined with Smart Bidding gives the algorithm more room to operate. That is true. It is also true that broad match in an account with thin conversion data, a weak audience signal, or no tROAS target will spend the budget on queries the business does not want.
Filter your keywords by match type and calculate the percentage of total spend sitting on broad. The thresholds I use:
The fix is rarely to eliminate broad. It is to cap broad with a tROAS target the business can actually achieve, or to move the broad spend into a separate campaign with its own budget so it cannot cannibalize phrase and exact performance.
Every account's campaign structure tells you what the person who built it was optimizing for. An account built for reporting has one campaign per service with descriptive names. An account built for performance has campaigns segmented by match type, intent, or margin. An account built in a hurry has a single Search campaign, one Shopping campaign, one PMax campaign, and no segmentation between them.
List every active campaign and capture the following for each:
The most common structural problem is overlap. A PMax campaign pulling clicks from a Search campaign on the same query, a Shopping campaign competing with PMax on the same SKUs, or brand search traffic getting credit for conversions non-brand actually drove. When campaigns overlap, reporting lies about which campaign is working.
The final structural check is interpretive, not mechanical. Most accounts generate weekly or monthly reports that show ROAS, CPA, conversion volume, and impression share. These numbers are real. They are also incomplete, and incomplete numbers lie more effectively than wrong numbers because they feel authoritative.
Four adjustments to how you read a report:
A report that does not address those four things is not wrong. It is just not enough to make a decision with.
The framework
Log in as admin. Verify the access level is Admin, not Standard or Read-only. If the agency controls the account under their MCC, request admin access in writing before continuing.
Tools, Conversions. Confirm each primary action has recent data, is marked Primary, uses a sane attribution model, and is not double-firing through GTM and Google tag simultaneously.
Insights, Search terms. 90-day range, sorted by cost. Read the top 30 queries. Flag anything that is not a real buyer. Build a negative keyword list in batches.
Filter keywords by match type. Calculate broad match percentage of total spend. Above 40 percent without a tROAS cap is a structural problem. Above 60 percent is a leak.
List every campaign, bid strategy, budget, and conversion goal. Find the overlaps. Flag any campaign where brand and non-brand are mixed, or where PMax and Shopping target the same SKUs.
Open each PMax campaign. Confirm asset groups are segmented by category or margin tier, audience signals are populated, and campaign conversion goals reflect the business outcome rather than a proxy event.
No. If you own the business, you own the Google Ads account. Log in at ads.google.com with the email that receives billing. If the agency set up the account under their own MCC and never granted you admin access, that is itself a finding. Request admin access in writing before the audit begins.
Open the Search Terms report, sort by cost descending, and read the top 30 queries. In most accounts, between five and twelve of those are queries the business should not be paying for. That is the fastest money you will find. It takes under fifteen minutes and usually surfaces recoverable spend worth multiples of a guide.
A full 90-minute audit every quarter. A Search Terms report review every two weeks. A conversion tracking verification after any website change, tag manager update, or Shopify theme edit. The audit cadence matters less than the tracking check, because a broken conversion event compounds daily.
Skip the PMax step and spend that time on match types and negative keywords. Standard Search and Shopping accounts have the same structural failure modes as PMax, just more visible. The audit sequence works identically. PMax is one of six checks, not the whole audit.
When the guide tells you what to look for but not what to do first. When the account runs across multiple markets, multiple conversion events, or more than three campaign types at once. When you find problems but cannot decide which to fix in what order. The Conversion Second Opinion exists specifically for that decision.
An audit is not a fix. An audit is a diagnosis. The value of running this sequence is not that you will leave with a better-performing account. You will leave with an accurate picture of what is happening inside the account, which is something most operators have never had. Decisions made from that picture will be better than decisions made from the dashboard.
The single highest-leverage finding in almost every account is conversion tracking. Fix tracking first. Everything downstream, bid strategy, match types, PMax, campaign structure, runs on the conversion signal. If the signal is wrong, every downstream fix is applied to the wrong target. If the signal is right, most other problems become visible and solvable without expert help.
When the audit surfaces more work than you can execute yourself, or when the problems interact in ways a framework cannot resolve, that is the point to bring in a practitioner. Stan Consulting offers Google Ads PPC management once the diagnostic is complete, and the Conversion Second Opinion is the entry point for anyone who wants the findings before the management.
Related: the full marketing guides collection covers Shopify, conversion, strategy, and agency management.
The engagement format
$5,000. One engagement. Diagnosis, build, and fix. No retainer after.
See the Revenue Sprint