Across more than 1,500 Google Ads accounts audited, the same waste sources appear in the same places. Over 70% of accounts have missing or broken conversion tracking. More than half lack any negative keyword lists. Most have bidding strategies set to modes the account is not yet eligible to use effectively. This 45-point Google Ads audit checklist walks through the ten areas where budget disappears without converting — in the order that surfaces the highest-value problems first.
Quick reference — what this checklist covers:
- 10 audit categories, 45 checkpoints total
- Estimated time: 90–120 minutes for a thorough review
- Most accounts: 3–5 critical issues, 8–12 moderate improvements
- Typical recoverable waste found: 25–40% of monthly spend
- Most common single finding: conversion tracking errors (affects over 70% of audited accounts)
Why Most Google Ads Accounts Have 30% or More in Wasted Spend
Google Ads is not a set-and-forget platform. Campaign settings degrade over time as your business changes, Google updates defaults, and competitor activity shifts the auction landscape. An account that was well-structured two years ago will have accumulated inefficiencies that are not visible from the dashboard — they only appear when you look at specific settings, search term reports, and conversion data together.
The 30% waste figure comes from accounts where the fundamentals are partially in place but not maintained. Accounts where basics were never set up — no conversion tracking, no negative keywords, generic broad match — can have 50–60% wasted spend from day one.
This checklist is designed to surface those issues systematically. Work through each category in order: conversion tracking issues make every other metric unreliable, so they need to be confirmed first before drawing conclusions from cost and performance data.
Category 1: Account Structure (5 Points)
Account structure problems are rarely the highest-cost issue, but they create downstream problems with reporting, bidding, and optimisation.
1.1 — Campaigns are segmented by objective, not by product or service. Each campaign should have one clear conversion objective. Mixing brand keywords, competitor keywords, and generic keywords in the same campaign creates bidding conflicts and obscures what is actually driving results.
1.2 — Ad groups contain tightly themed keyword sets. Ad groups with more than 15–20 keywords are almost always too broad. Review each ad group: if the keywords could not all plausibly trigger the same ad and land on the same page, split them.
1.3 — No more than 3–5 active text variants per ad group. Running too many ads dilutes impressions and makes it harder to reach statistical significance for testing. Two to three RSAs per ad group is the practical ceiling for most accounts.
1.4 — Separate campaigns exist for brand vs non-brand. Brand and non-brand keywords should never share a campaign or budget. Brand CPCs are 5–10× lower and conversion rates are 3–5× higher — mixing them distorts every performance metric.
1.5 — Correct campaign types are used for each objective. Search for intent-based queries, Performance Max for signals-based automation, Shopping for product feeds. Demand Gen and Display for awareness only if the funnel warrants it. Using Shopping campaigns for service businesses or Search campaigns for e-commerce product catalogues is a structural mismatch.
Category 2: Conversion Tracking (8 Points)
Broken conversion tracking is the single most costly issue in Google Ads. When the algorithm does not know what a conversion is, Smart Bidding optimises against noise. Identifying and fixing this is the highest-ROI action in most audits.
2.1 — At least one conversion action is marked “Primary” with a non-zero conversion count. Open Conversions in the Goals menu. If no action has “Primary” status with recent conversions, Smart Bidding has no signal to work with.
2.2 — Conversion actions match the actual business objective. Phone calls, form submissions, purchase completions — not page visits or session duration. Check the “Conversion source” column to confirm each action fires at the right point in the funnel.
2.3 — No duplicate conversions are inflating counts. If form submission fires both a Google Ads tag and a Google Analytics imported goal, every lead counts twice. Duplicate tracking overstates volume and causes Target CPA/ROAS to underbid.
2.4 — Tag coverage is confirmed in Google Tag Diagnostics or Tag Assistant. Run Tag Assistant on your conversion page. If the tag fires once on page load and again on a repeat visit without a new form submission, you have a misfire. If it never fires, you have a gap.
2.5 — Call tracking is active if phone leads are a significant conversion source. Most service businesses receive 40–70% of inbound enquiries by phone. If there is no call tracking, those conversions are invisible to the bidding algorithm.
2.6 — Enhanced conversions or consent-mode-compliant tracking is configured. Since GDPR/Privacy Act changes, cookie consent impacts data completeness. Enhanced conversions using hashed first-party data recover a meaningful portion of lost conversion signals — typically 10–25% of conversions in consent-restrictive environments.
2.7 — Attribution model is appropriate for the sales cycle. Data-driven attribution is the default and usually the best choice. Last-click is only appropriate for extremely short sales cycles (under 24 hours). If your sales cycle is longer than a week, time-decay or linear attribution will give more accurate cross-channel credit.
2.8 — Offline conversion import is set up for businesses with offline close rates. If your leads close by phone, in-person meeting, or after a multi-week sales process, offline conversion import feeds actual closed revenue back to the campaign. Without it, the algorithm optimises for lead volume, not lead quality — and those two things diverge significantly for most B2B and high-ticket service businesses.
Category 3: Keyword Strategy and Match Types (6 Points)
The keyword layer is where budget leaks quietly over time. Match types have changed significantly since 2021 — what was tight phrase match behaviour now operates closer to broad match in many cases.
3.1 — No active broad match keywords without audiences attached. Broad match without audience targeting or a robust negative list will pull in irrelevant queries at scale. If you are using broad match, confirm that it is paired with Customer Match or remarketing audiences and a solid negative list.
3.2 — The search terms report has been reviewed in the last 30 days. If the search terms report has not been reviewed, you do not know what queries are actually triggering your ads. Set a recurring 30-minute task to review it weekly or fortnightly.
3.3 — High-spend, zero-conversion keywords have been paused or removed. Any keyword with 50+ clicks and zero conversions over a 90-day window should be paused and reviewed. It is either poorly matched to intent or sending traffic to a non-converting page.
3.4 — The keyword list is not duplicating intent across match types. Running “physio Sydney” (exact), “physio Sydney” (phrase), and “physio Sydney” (broad) in the same campaign creates an internal auction where you bid against yourself. Consolidate to one match type per theme, or use single keyword ad groups if click volume warrants it.
3.5 — Long-tail keywords are present for consideration-stage queries. High-intent, low-volume keywords like “google ads management for ecommerce Sydney” have lower CPCs, higher conversion rates, and clearer commercial intent than head terms. Most accounts under-invest in the long tail.
3.6 — Competitor brand keywords are in a separate campaign with a separate budget. Competitor terms carry different CPCs, quality scores, and conversion rates than your own brand or generic terms. They need their own budget cap and their own assessment of ROI.
Category 4: Negative Keywords (4 Points)
Negative keywords are the fastest fix in most accounts. Adding 50–100 well-chosen negatives to an account that has none will often produce a measurable CPC reduction within two weeks.
4.1 — A negative keyword list exists at the account level. Universal exclusions — “free”, “DIY”, “jobs”, “course”, “how to”, “salary”, “template” (unless you sell templates) — should be in an account-level list shared across all campaigns.
4.2 — Industry-specific negatives are applied to relevant campaigns. A removals company should exclude “packing materials”, “cheap”, “lorry hire”. A dental clinic should exclude “free”, “university”, “NHS” (if private). A mortgage broker should exclude “calculator”, “comparison”, “broker salary”.
4.3 — Negative keywords are reviewed after each search terms audit. Every irrelevant query in the search terms report that spent more than $5 without converting is a candidate for a negative keyword. This is a maintenance task, not a one-time setup.
4.4 — Negative match types are correct. Negative exact prevents that exact term. Negative phrase prevents queries containing that phrase in sequence. Negative broad prevents queries containing all those words in any order. Using negative broad for product terms can block relevant queries unintentionally.
Category 5: Bidding Strategies (5 Points)
Smart Bidding is powerful when it has enough data and the correct targets. The most common error is using automated bidding before the account has enough conversion volume for the algorithm to work from.
5.1 — Smart Bidding campaigns have at least 30 conversions per month per campaign. Google’s own Smart Bidding documentation recommends 30 conversions per month as the minimum for Target CPA and Target ROAS. Below that threshold, the algorithm is learning from too few data points and will overpay or underpay in unpredictable ways.
5.2 — Target CPA and Target ROAS are set based on actual historical data, not aspirational targets. Setting Target CPA 50% below your actual CPA will cause the algorithm to reduce bids until impressions and clicks drop to near zero. Set targets at or slightly above current performance, then reduce gradually.
5.3 — Budget caps are not chronically limiting delivery. If a campaign shows “Limited by budget” status regularly, it is not getting the impressions available at that CPA target. Either increase the budget or increase the CPA target to reduce cost per click and extend reach.
5.4 — Maximise Conversions is not set on low-budget campaigns with no target. Maximise Conversions without a CPA target will spend the entire daily budget in the most expensive way possible. For campaigns under $50/day AUD with no historical conversion data, manual CPC or Enhanced CPC is a safer starting point.
5.5 — Portfolio bid strategies are used where multiple campaigns share the same conversion goal. If you have three Search campaigns all targeting the same lead type, a Portfolio Target CPA strategy lets the algorithm allocate budget fluidly across all three rather than hitting each campaign’s target in isolation.
Category 6: Ad Copy and Assets (5 Points)
Ad strength scores in Google’s UI are proxies for asset coverage, not for conversion quality. An “Excellent” ad strength rating does not mean the ad is converting. Focus on message relevance and offer clarity over asset count.
6.1 — Each RSA has at least one headline and one description that mirrors the keyword intent. If the search is “google ads agency sydney”, one headline should contain those words or close variants. Pinning is appropriate for brand and legal accuracy, but do not pin more than two headlines or the dynamic benefits of RSA are lost.
6.2 — A clear, specific offer or value proposition is in every ad. “We are a Google Ads agency” is not an offer. “Managed campaigns from $900/month” or “Average $4.60 CPC for tradies in Melbourne” is an offer. Specificity improves CTR and pre-qualifies clicks.
6.3 — All relevant extensions (assets) are active: sitelinks, callouts, call, lead form. Every unused asset type is a missed opportunity to increase ad real estate. At minimum: 4 sitelinks, 4 callout extensions, call extension (if you accept phone calls), and location extension if you have a physical address.
6.4 — Ad copy does not contain superlatives or claims that require substantiation. “Best Google Ads agency in Australia” may violate Google’s ad policies and Australian Consumer Law. “Google Ads agency specialising in small business” is accurate and compliant.
6.5 — Display URL paths are used with keywords, not left blank. Display URL paths (the two optional fields after the domain) are free keyword placement. A display URL of adstralis.agency/Google-Ads/Sydney increases relevance and CTR at no extra cost.
Category 7: Audiences (3 Points)
Audience layers are the most underused feature in most Search campaigns. They add zero cost until you apply bid adjustments, but they give you performance data that is otherwise invisible.
7.1 — Remarketing lists are applied as observation audiences to all Search campaigns. Even if you are not adjusting bids for remarketing audiences, having them applied in observation mode reveals conversion rate differences between new and returning users — data that informs later bid adjustments.
7.2 — Customer Match is active if a customer email list is available. Uploading a customer list to Customer Match and applying it as an observation or targeting audience lets you exclude existing customers (reducing wasted spend) or bid more aggressively for high-value prospect segments.
7.3 — Audience bid adjustments are set based on observed performance differences. If your remarketing audiences convert at 3× the rate of new users, a +50–100% bid adjustment is justified. If mobile users convert at half the rate of desktop users, a -25% mobile bid adjustment reduces waste without cutting reach significantly.
Category 8: Performance Max Settings (4 Points)
Performance Max is increasingly where Google allocates budget by default. Most accounts have it running with factory settings — which means broad reach and limited control over where spend goes.
8.1 — Asset groups are segmented by product or service type, not lumped together. A PMax campaign with one asset group covering your entire catalogue will favour the assets with the broadest appeal, which is often not the most profitable product line. Separate asset groups by margin or product category.
8.2 — Brand exclusions are applied at campaign level. Without brand exclusions, PMax will take credit for branded search traffic that would have converted anyway at a lower CPC via a standard Search campaign. Apply brand name exclusions to your PMax campaigns.
8.3 — URL expansion is either controlled or switched off for landing-page-sensitive accounts. URL expansion lets Google send traffic to any page on your site it deems relevant. For accounts with specific landing pages designed to convert, or for lead generation where a specific form page must receive traffic, switch URL expansion off or enable page feeds to control it.
8.4 — Search themes are set to guide asset group intent. Adding search themes to PMax asset groups signals what queries that group should target. Without them, the algorithm relies entirely on landing page content and past campaign data — which can produce mismatches in the early weeks of a new campaign.
Category 9: Shopping and Feed Quality (3 Points)
Applies to e-commerce accounts running Shopping campaigns or PMax with product feeds.
9.1 — Feed has no disapproved products and no “Limited performance” warnings. Check Merchant Center for feed errors. Disapproved products are excluded from all ad auctions. “Limited” status usually means missing GTINs, image quality issues, or price mismatches between the feed and landing page.
9.2 — Product titles contain the primary search keyword in the first 70 characters. Product titles are Google Shopping’s primary relevance signal. “Red Women’s Running Shoes Nike Pegasus 40 Wide” outperforms “Nike Pegasus 40” because it front-loads the search terms most likely to trigger impressions.
9.3 — Custom labels are used for margin segmentation. If you are running PMax or Smart Shopping, custom labels let you tell the algorithm which products have higher margins (and thus deserve more aggressive bidding). Without labels, the algorithm bids the same for a 60% margin product and a 15% margin product.
Category 10: Reporting and Attribution (2 Points)
10.1 — A weekly performance dashboard exists with the three KPIs that matter: cost, conversions, and cost per conversion. Dashboard overload kills optimisation. A single view showing weekly trends for spend, conversion volume, and CPA is more actionable than a 30-metric report. If you are not looking at it weekly, it does not exist.
10.2 — Search impression share and lost IS (budget vs rank) are reviewed monthly. Impression share lost to budget tells you if you need more spend. Impression share lost to rank tells you if you need better Quality Scores or higher bids. These two metrics together diagnose whether underperformance is a budget problem or a quality problem — which require completely different fixes.
What to Do After the Audit: Priority Order for Fixes
Not all findings are equal. After completing the 45 points, categorise findings by impact:
Fix immediately (account integrity):
- Broken or missing conversion tracking
- Campaigns with no conversions in 90 days and no paused status
- Negative match type errors causing relevant traffic to be blocked
Fix within one week (significant waste):
- Broad match keywords without negative lists
- Target CPA/ROAS set below historical performance
- No negative keyword list at account level
Fix within one month (optimisation):
- Ad group restructuring
- Asset coverage gaps
- Performance Max brand exclusions
- Audience observation layers
Internal links for reference: the enhanced conversions setup guide covers point 2.6 in detail. The conversion tracking guide covers offline import and call tracking. For Performance Max specifically, the Performance Max guide explains the brand exclusion and asset group settings in points 8.1–8.4.
If you have worked through this checklist and identified more than five critical issues, a structured account rebuild is likely more efficient than incremental fixes. Get in touch and we will review the specific issues your account has and walk through the priority order for your situation.
Frequently Asked Questions
How often should you audit a Google Ads account?
A full 45-point audit should be done quarterly for active accounts. A lighter 15-point review of conversion tracking, search terms, and wasted spend should happen monthly. Accounts with more than $5,000 AUD/month in spend warrant monthly full reviews.
What is the first thing to check in a Google Ads audit?
Conversion tracking. Every other metric is meaningless if the conversion data is broken. Before reviewing CPA, ROAS, CTR, or CPC, confirm that the conversion actions marked “Primary” are firing correctly on the right pages with the right conversion windows.
How do you know if your Google Ads account is performing well?
Benchmarks vary by industry, but broadly: for service businesses in Australia, a CPA under $80–120 AUD for low-ticket services and under $200–400 AUD for high-ticket services is a reasonable starting point. For e-commerce, a ROAS above 3.0 across all channels (not just Google) generally indicates a profitable mix. Your own historical CPA and ROAS trends are more meaningful than industry benchmarks for any specific account.
Can you audit a Google Ads account without admin access?
You can audit at a high level with read-only access to the account. However, conversion tag verification requires Tag Assistant on the live site (no account access needed), and Merchant Center feed errors require Merchant Center access. For a full audit, read-only Google Ads access plus standard site access is sufficient.
What is the most common Google Ads mistake found in audits?
In accounts we have reviewed, missing or broken conversion tracking appears in over 70% of cases. It is the most costly error because it silently corrupts every bidding decision the algorithm makes. The second most common issue is the absence of negative keyword lists — present in more than 50% of audited accounts.
What does a Google Ads audit cost if done by an agency?
A professional audit from a specialist agency in Australia typically costs $500–$1,500 AUD depending on account complexity and depth of analysis. Some agencies offer free audits as a sales tool — these are typically surface-level reviews that identify obvious issues but rarely surface the structural problems in category 2, 5, and 8 above.