GOOGLE ADS AUDIT: THE SEQUENCE I FOLLOW

Most people run a google ads audit as a checklist and hand the client back forty findings of equal weight. I have inherited plenty of those documents, and they are almost useless, because a broken conversion tag and a slightly weak headline sit on the same line with the same tick box. The value of an audit is not in how many things you noticed. It is in the sequence you noticed them in, and in what you inferred from each one. Run the checks in the wrong order and you draw confident conclusions from data that was lying to you the whole time. This article is the order I actually follow, why tracking and structure come before everything else, and how I turn the raw findings into a ranked action plan instead of a shopping list.

The discipline behind the whole thing is simple: an audit is a diagnosis, not a treatment. I resist the urge to fix anything while I am still looking, because the moment you start editing you stop seeing the account as it really is. I read, I note, I infer, and only at the very end do I decide what to touch first. What separates a useful audit from a spreadsheet nobody acts on is that last step, and it depends entirely on getting the first steps right.

Before I go further, a word on what an audit is for. Auditing your google account is not an exam you pass or fail; it is a way of building an accurate picture of reality so your next decisions are grounded rather than guessed. A good ads account audit is fundamentally a data driven exercise: I do not trade in opinions about whether a campaign "feels" tired, I trade in what the numbers say once I have confirmed the numbers can be trusted. That distinction runs through everything below. Every finding I keep has to be backed by evidence I can point to, and every priority I assign has to be defensible with a figure. The rest of this piece is how I get from an account I have never seen to a short, ordered list of things worth doing, in the order that actually protects me from fooling myself.

How to audit a google ads account: the order I follow

The reason sequence matters is dependency. Every layer of an account inherits its trustworthiness from the layer beneath it, so if you audit top-down you keep validating things built on foundations you have not checked. I go bottom-up instead. When I learn how to audit a google ads account for a new client, the first question is never "how are the ads performing" but "can I believe the numbers that would tell me." Only once I trust the measurement do I let performance data influence a single conclusion.

Concretely, I move through five layers in this order: measurement, then account-level settings and structure, then campaign settings and bid strategy, then search terms and query-level waste, then the ads and assets themselves. Notice that the flashy stuff, ad copy and creative, comes last. That is deliberate. A brilliant headline on a campaign optimizing toward a double-counted conversion is a rounding error. A well-structured account feeding clean signals into the auction is where the real money is won or lost. The sequence forces me to spend my attention where it compounds, and it stops me drawing inferences from data I have not yet earned the right to trust.

This ordering holds whether the account is tiny or enterprise-scale, and whether it runs pure paid search or a mix of everything. Google ads campaigns have proliferated in type over the years, Search, Shopping, Performance Max, Demand Gen, Video, and it is tempting to audit each type with its own bespoke ritual. I resist that. The dependency chain is the same underneath all of them: measurement has to be sound before structure means anything, and structure has to be sound before performance means anything. What changes by campaign type is the detail I inspect within each layer, not the layer order. So I keep one spine for every account and adapt the questions, rather than inventing a new process for each surface and losing the discipline that makes the audit trustworthy in the first place.

There is a Google Ads Community write-up from a manager auditing a large enterprise account that maps almost exactly onto this order, moving from account settings to tracking to campaigns to ads, and you can read their account audit structure they followed. What that account teaches, and what my own accounts keep confirming, is that the expensive mistakes hide in the bottom two layers while everyone's attention is on the top.

What I check first: tracking

Tracking is the first thing I open, every single time, before I have looked at a single performance number. The logic is a chain: if tracking is broken the data is wrong, if the data is wrong the insights are wrong, and if the insights are wrong every decision downstream is wrong while you keep spending. Auditing performance before verifying measurement is auditing fiction. So I earn the right to trust the account before I let it tell me anything.

Conversion actions and what counts as primary

My first move is to list every conversion action and check which are set as Primary, because those are the ones Smart Bidding actually optimizes toward. The classic failure I keep finding: a lead-gen account with downloads, outbound clicks, and other micro-conversions all flagged as Primary alongside the actual lead. The campaigns were dutifully optimizing toward the cheap, easy actions instead of qualified leads, and nobody had noticed because the conversion count looked healthy. When I audit the goals feeding conversion tracking, I am really auditing what the algorithm has been told to want. Get that wrong and every clever bid decision underneath is aimed at the wrong target.

Double-counting and the numbers that look too good

The second failure is subtler and more expensive: a conversion firing twice on the thank-you page. One real lead, two recorded conversions. Cost per lead looks better than reality, the dashboard looks strong, and the account appears to be outperforming when it is quietly misreporting. I always cross-check Google Ads conversions against the source of truth, the CRM or the back end, and when the two disagree that gap becomes the single highest-priority finding in the whole audit. Everything else can wait; a measurement that lies corrupts the entire diagnosis.

Only when I have confirmed conversions are complete, correctly attributed, and counted once do I let any performance figure influence a conclusion. This is also why, on any inherited account, tracking verification is the first block of my handover audit for accounts I take over: the previous manager's reported CPL means nothing until I have proven the meter was reading true.

Cross-checking the platform against a second source

My last tracking step is triangulation. I never take Google's conversion count as gospel on its own; I line it up against a second source, usually Google Analytics, and see whether the two tell the same story. When the platform reports a conversion rate that Analytics cannot corroborate, or when session and click numbers diverge in ways auto-tagging should have prevented, that discrepancy is itself a finding. I also sanity-check the click through rate against what the account structure would predict: an oddly high click through rate on a broad, low-intent campaign is often a sign of a measurement or matching problem rather than a genuine win. None of these figures is trustworthy in isolation. Read together, they either reinforce each other or expose the crack, and finding the crack early is the whole reason tracking sits first in the sequence.

What I look at second: account and campaign structure

With measurement trusted, structure is where I spend the bulk of the audit, because structure is what shapes every auction the account enters. I read it as evidence, not as a checklist, and each thing I find lets me infer something about how the account was really being run.

Settings, bid strategy, and goal alignment

I download campaign-level data and read the settings as a set: budget, campaign type, bid strategy, networks, locations. The inference I am chasing is whether the bidding strategies in place match the business objective. A lead-gen account running Maximize Clicks is a tell that someone was chasing volume over qualified demand. Search campaigns with Display expansion silently switched on is a tell that reach was prioritized over lead quality. One global campaign spending most of its budget outside the target geography, which I have found more than once, is a tell that nobody had read the location report in a long time. None of these are just boxes to tick; each is a clue about the account's history and where the waste is concentrated.

Below the campaign layer I read the ad groups, because their granularity is another tell. Sprawling ad groups stuffed with loosely related keywords usually mean weak intent alignment, and weak intent alignment shows up later as a depressed quality score and a higher cost per click than the account should be paying. When I see a poor quality score sitting next to bloated ad groups, I do not treat them as two separate findings; I treat the structure as the cause and the quality score as the symptom, and that inference tells me which one to fix. I also flag any Display campaigns bundled in where a lead-gen objective sits, since prospecting display and high-intent search have no business sharing a budget or a bid strategy, and mixing them muddies every metric that follows.

Where structure hides missed demand

The other half of the structure audit is what is not there. An account living entirely on brand-exact keywords is capturing existing demand and nothing else, missing everyone searching generic and competitor terms who does not yet know the brand. That is not a broken setting, so a checklist never flags it, but it is often the biggest growth lever in the account. I also check whether the account is leaving free google ads visibility on the table, for instance unclaimed Merchant Center free listings on an ecommerce account, because paid coverage without the organic surfaces around it is a partial strategy. Reading structure this way, as evidence of both waste and absence, is what separates an audit that finds problems from one that finds opportunities.

Search terms as the proof of structural leaks

The search terms report is where structural weakness becomes a number. I pull a long window, often a year or more, and quantify how much spend went to irrelevant search terms that were never going to convert. On accounts with weak match-type control and thin negative keywords lists, I routinely find a large share of budget leaking to queries that had no commercial intent at all. That figure is not a vague concern; it is a hard, defensible number I can put on the action plan with a euro sign attached, which is exactly what turns a finding into a priority. The search terms report also does double duty as an expansion tool: the same document that exposes waste surfaces the high-intent queries the account is not yet targeting, so one careful read gives me both the negatives to add and the keywords to build out.

The last layer: ads and landing pages

Only now, with measurement and structure understood, do I look at the ads themselves, and I always look at the ad and landing page together rather than in isolation. An ad promising one thing and a page delivering another is a leak no amount of bidding cleverness will fix, so I read them as a pair: does the message continue from click to page, does the page ask for the action the campaign is optimizing toward, does it load and convert. Weak ad strength gets noted here too, but I keep it in perspective. Rewriting headlines feels productive, yet on most accounts the campaigns and ads at the top of the interface matter far less than the tracking and structure underneath, and my ranking reflects that.

Landing pages deserve their own line in the audit because they sit at the boundary between what I control in the account and what the client controls on the site. When conversion rate is poor despite healthy traffic and clean intent, the landing pages are usually where the story ends, and flagging that clearly, with evidence, is often the single most valuable thing an audit hands back, even though the fix lives outside Google Ads.

Turning the audit into priorities, not a shopping list

This is the step that most audits skip and the one clients actually pay for. Once the reading is done, I have a page of findings, and a page of findings is worthless until it is ranked. The audit itself is not the time to make changes, unless something is so critical it must be fixed on sight; the reading and the fixing are separate acts, and collapsing them is how audits turn into panic instead of strategy.

I rank every finding on two axes: how much money it moves, and how hard it is to fix. A double-counted conversion is high-impact and low-effort, so it goes to the top and gets fixed today. Thirty percent of spend on irrelevant search terms is high-impact and medium-effort, so it becomes the first structural project. A batch of average-strength ads is real but low-impact next to the others, so it waits, no matter how satisfying it would be to rewrite them. I keep the raw findings in a structured audit checklist template so nothing gets lost, but the ranking is what I actually work from. The filter I run every finding through is one question: if I fix this, does cost per conversion or return on ad spend actually move? If the honest answer is "not much," it drops down the list regardless of how obviously "wrong" it looks. That single question is what keeps the plan focused on leverage instead of tidiness.

What the client receives, then, is not forty equal ticks but a short ordered list: fix this today, build this next, revisit these later. Each item carries the inference behind it and, wherever possible, the number it moves. That is the whole point of running the checks in sequence, because trusting the tracking is what lets me attach real figures to structural findings, and real figures are what let me rank. From there the work flows straight into an ongoing rhythm, and the audit's ranked list becomes the opening agenda for my weekly optimization routine rather than a document that gets read once and filed. An audit that ends in priorities gets acted on. An audit that ends in a shopping list gets admired and ignored.