How to audit a Google Ads account: search terms, settings, conversion tracking, and budget allocation

A Google Ads audit is a mechanical pass through tracking, queries, settings, and budgets so you can trust the numbers before you decide what to scale, fix, or shut down. This sequence starts with conversion tracking, then search terms, then settings that quietly move spend, and ends with a budget and impression share read; deeper tracking implementation belongs in the dedicated conversion tracking guide.

Start with conversion tracking status and why it leads the audit

Open Google Ads and go to Goals > Summary. This is where Google surfaces conversion action status, which is the first gate in any audit because every ROAS, CPA, or CPC decision depends on whether conversions are being recorded correctly.

Per Google’s documentation, each conversion action can currently be in 1 of 5 statuses: Active, Needs attention, Misconfigured, Awaiting conversions, or Removed. Active means the action is set up and has recent activity, Needs attention flags potential issues, Misconfigured means conversions have stopped recording entirely due to a setup error or broken tag, Awaiting conversions means none have recorded in the last 7 days, and Removed indicates the action is no longer in use.

For the audit, list every conversion action that should represent a business outcome (purchases, qualified leads, key product actions) and confirm they are Active. Any Misconfigured or Needs attention status on a core action means you cannot trust performance metrics for the affected period and must fix tracking before interpreting campaigns.

Awaiting conversions on a minor or experimental action may be acceptable, but Awaiting conversions on your primary revenue or lead action is a red flag for volume or implementation. Removed actions should not be referenced in current reporting or bidding; if they are, you need to update strategies and columns to use live actions instead.

This is why an audit starts here: if conversions are missing or misclassified, every later decision about bids, budgets, or keywords is based on distorted data.

Check conversion action configuration: primary, counting, and windows

Still in Goals > Summary, click into each conversion action that should drive optimization. You are checking three things: whether it is Primary or Secondary, whether it counts Every or One conversion, and what the conversion window is.

Google’s help pages state that each action can be marked Primary or Secondary, and only Primary actions feed the Conversions column that bidding strategies use. Secondary actions are for observation only. In an audit, any soft actions (newsletter signups, page views) that should not steer bids should be Secondary, and your core revenue or lead actions should be Primary so that the Conversions column reflects business value.

Next, check the counting setting. Google documents two options: Every, which counts each conversion per click and is suited to purchases, and One, which counts a single conversion per click and is suited to leads. Every is the default for website and import actions, and One is the default for calls and app installs. If a lead form action is set to Every, you may be overcounting repeat submissions from the same click; if an ecommerce purchase is set to One, you may be undercounting multiple orders from a single click.

Then review conversion windows. For Search, Shopping, and Display, click-through conversion windows are configurable from 1 to 90 days, and engaged-view windows from 1 to 30 days, according to Google’s documentation. Window changes are not retroactive, so an audit should record the current windows and interpret historical performance in that context rather than expecting past data to update.

For each key action, document: Primary or Secondary, Every or One, and the window length. Align these with your business model before you move on, because misaligned counting or windows can make ROAS and CPA look better or worse than they are.

Audit the search terms report and its privacy limits

Once conversions are trustworthy, move to query quality. Go to a Search campaign or the account-level view, then Reports > Search terms. This report shows the actual search terms that triggered your ads, which is where you find wasted spend and new keyword opportunities.

Google’s help pages specify that the search terms report shows only terms used by a significant number of people, and low-volume queries are omitted for privacy. As a result, the spend in the visible search terms will not sum to your total campaign spend. This corrects the misconception that the search terms report shows everything the account matched; it never does, so you cannot reconcile it line by line to total cost.

For the audit, filter the search terms report by cost and conversions over your review window. Identify high-cost, zero-conversion terms that are clearly irrelevant and mark them for negatives. When you add negatives directly from this report, Google notes that they default to exact match, so you need to be deliberate if you want broader blocking, which is covered in the separate negative keyword strategy guide.

Also scan for high-converting terms that are not present as exact or phrase keywords. These are candidates to promote into keywords so you can control bids and match types more precisely. Because of the privacy threshold, you will miss some low-volume waste and some long-tail winners, but the visible terms still capture a large share of spend and are the main lever for cleaning up query quality.

Identify silent-waste campaign settings: networks, locations, rotation

Next, audit campaign-level settings that quietly move budget. Open each Search campaign, click Settings, and review Networks, Locations, and Ad rotation.

Under Networks, Google’s documentation states that Search partners are included by default on new Search campaigns, with an opt-out toggle labeled “Include Google search partners.” This corrects the misconception that default settings are neutral; they encode Google’s preferences. In an audit, note whether Search partners is on and compare performance by segmenting the campaign by network before deciding to keep or disable it.

If the campaign uses Display Expansion, Google describes this as serving on Display only when it predicts unutilized Search budget, with an upcoming migration to channel prioritization and budget allocation starting February 2027. For now, your audit should flag any Search campaign with Display Expansion enabled, because it can move budget into Display inventory that behaves differently from Search, which complicates ROAS and CPC reads.

Under Locations, open Location options. Google’s help pages explain that the default “Presence or interest” includes people who have shown interest in the targeted locations, while “Presence” restricts to people likely to be in them. Exclusions always use Presence. This is another place where defaults are not neutral; if you intend to target only users physically in your regions, switch to Presence and document the change.

For Ad rotation, Google documents two settings: Optimize and Do not optimize. Under Smart Bidding, Google automatically uses Optimize regardless of the setting. That means if your campaigns use Smart Bidding, the rotation control is effectively overridden, which again shows that defaults follow Google’s optimization logic rather than advertiser testing preferences. In an audit, simply confirm that this behavior matches your bidding strategy expectations.

Handle auto-apply recommendations and Optimization Score

Now check how much control Google has over changes. Go to Recommendations, then the Auto-apply section. Google states that auto-apply recommendations are opt-in by bundle, such as “Maintain your ads” and “Grow your business,” or by individual type, and that any auto-applied change appears in the History tab.

Google also documents that the Maintain Your Ads set never increases or changes budget. This is relevant to the misconception that auto-apply recommendations are safe to leave on by default. They are opt-in for a reason, and your audit should list which bundles or individual auto-applies are enabled, then cross-check the History tab to see what has been changed automatically.

Optimization Score appears at the top of the Recommendations page and ranges from 0 to 100%. Google describes it as an estimate of how well the account is set up to perform, not a measure of how it is performing. Both applying and dismissing recommendations can raise the score, and Google publishes no formula behind it.

This corrects the misconception that a high Optimization Score means the account is performing well. In TF’s practice, an audit exists to answer operating questions, not to raise Google’s scores, so you should read Optimization Score as Google noise with occasional insight, and audit toward ROAS, CPC, and deliverability instead of chasing 100%.

Read budget allocation, Limited by budget, and impression share

With tracking, queries, and settings checked, move to budget allocation. At the campaign level, look at the Status column for “Limited by budget” flags, and add impression share columns: Search impression share, Search lost IS (budget), and Search lost IS (rank).

Google documents that “Limited by budget” appears when a campaign underperforms due to budget or when a Maximize Clicks strategy could buy more traffic. The fix flow offers a recommended daily budget based on recent performance, current budget, keywords, and targeting. This corrects the misconception that a Limited by budget flag always means underfunding; a non-converting or low-ROAS campaign limited by budget can be capped correctly.

Google’s help pages explain that Search lost IS (budget) is available only at the campaign level, and that Search lost IS (rank) measures losses due to poor Ad Rank. Together, these two buckets split budget problems from quality problems. If a converting campaign has high lost IS (budget) and reasonable ROAS, you are leaving volume on the table. If lost IS (rank) is high, the constraint is bids, ad relevance, or landing page quality, not budget.

Google also notes that impression share metrics update within 1 to 2 days, and account statistics can be delayed up to 3 hours. For an audit, avoid reading same-day impression share; use a window where data has settled so that your budget and rank conclusions are stable.

If your audit finds misallocated spend but you need a structured way to size budgets, that methodology sits in a separate budget planning guide, while here you are only classifying campaigns as overfunded, underfunded, or correctly capped based on Limited by budget and the lost-IS buckets.

Make TF-style keep, fix, or kill decisions from the metrics

After the mechanical checks, the audit becomes a read on performance. TF’s operating lens is ROAS, CPC, and deliverability, with search impression share as the biggest validator of whether spend is allocated correctly.

For each campaign, ask: does it hold ROAS at a level that aligns with your economics, are CPCs reasonable for that return, and is deliverability (impression share) constrained by budget or rank. Converting campaigns with solid ROAS and high lost IS (budget) are “keep and consider scaling” candidates, as long as you accept the unit economics. Converting campaigns with high lost IS (rank) move into “fix” territory, where bids, ad relevance, or landing pages need work before more spend.

Campaigns that do not convert or that deliver poor ROAS despite adequate impression share and no Limited by budget flag fall into “kill or radically reduce” territory. This is where the misconception about Limited by budget is most dangerous: a non-converting campaign that is Limited by budget is not underfunded, it is correctly constrained until it proves it can convert.

TF’s practice is to treat Quality Score and Optimization Score as secondary signals at best. The primary decision is always whether the campaign delivers acceptable ROAS at sustainable CPCs, and whether impression share data shows room to grow without pushing past the inflection point where more impression share costs more ROAS than it returns. For recurring audits, you can automate parts of this reporting with scripts, which are covered in a separate automation guide.

In TF’s practice, an audit exists to answer operating questions, not to raise Google’s scores. The diagnostic lens is always ROAS, CPC, and deliverability, with search impression share as the biggest validator of whether spend is allocated right, and Quality Score and Optimization Score are treated as secondary signals with occasional insight rather than primary goals.