You’ve been running Google Ads for a while. The spend is real, but you’re not confident the account is set up correctly. Maybe results have been inconsistent. Maybe someone told you the account “looks fine” but you still have doubts. Maybe you just changed who manages it and want to know what you’re inheriting.
An audit is the structured answer to that uncertainty.
What a Real Audit Is
A Google Ads audit is a systematic review of every layer that determines whether an account is likely to perform well or waste money. That includes the measurement layer (is the account tracking the right things accurately), the structural layer (are campaigns organized in a way that matches how the business actually works), and the execution layer (are the ongoing decisions being made well).
This is different from what Google’s interface calls an audit. Google’s optimization score and built-in recommendations are designed to move accounts toward more automation and more spend. They are not a diagnostic. A real audit looks at whether the account matches how your business makes money, not whether you’re following Google’s preferred settings.
For a typical Shopify store running Shopping, Performance Max, and Search, a thorough audit takes between three and five hours of focused work. What comes out is a prioritized list of specific problems, each with a clear explanation of what it costs and what fixing it involves. Not a list of generic suggestions.
The First Layer: Conversion Tracking
Before I look at a single campaign, I verify that the account is measuring the right things correctly. This is not a formality. Across the accounts I have reviewed, broken or misconfigured conversion tracking is the single most common finding, and it corrupts every analysis and decision downstream.
The specific things I check:
Whether purchase conversions are actually being tracked. Some accounts have no purchase conversion action at all. They are measuring softer signals, like “added to cart” or “visited a product page,” as their primary conversions. The bidding algorithm optimizes for whatever it is told is a success. If you tell it that adding to cart is a success, it finds people who add to cart. Not necessarily people who buy.
Whether conversions are being counted more than once per order. This is a silent problem that I find regularly. It happens when a store has Google Ads conversion tracking set up through Google Tag Manager and also has the Google Ads tag placed directly on the thank-you page, both firing on the same purchase. Every order triggers both. The account reports 1.8 or 2.3 conversions per actual order. ROAS looks great. Nobody notices because the absolute numbers seem plausible.
Whether order values are being passed through correctly. If the conversion tracking fires but does not include the revenue amount, every order looks the same size to the algorithm. A 15-dollar order and a 1,500-dollar order are treated identically. The campaigns optimize toward volume, not revenue.
Whether a recent site change broke tracking silently. I compare the conversion data timeline against the site change history. A Shopify theme update, a new checkout app, a change to the thank-you page URL, or a GTM container publish can all stop conversion tracking without any error messages or alerts. The store keeps getting orders. Google Ads stops seeing them.
If I find a tracking problem in the first layer, that takes priority over everything else. There is no point diagnosing campaign structure when the measurement is wrong.
The Second Layer: Campaign Structure
Once I am confident the measurement is correct, I look at how the account is organized. Structure determines whether the automation has the right boundaries to make good decisions.
The most common structural problem I find in Shopify stores is a single Performance Max campaign running the full product catalog with no brand exclusions, no product segmentation, and no meaningful audience signals. Everything is in one bucket. Google decides where to spend, and it typically decides to spend heavily on branded search (people already searching your store name) and retargeting (people who have already visited your site). These placements look great in reports because existing customers convert well. But those customers were coming back regardless of the ads. The store paid for something that was going to happen anyway.
What a properly structured account looks like instead: a clear separation between branded and non-branded activity. A branded campaign protects your search terms from competitors, but its results are tracked separately so they do not inflate the metrics for new customer acquisition. Non-branded campaigns are assessed on their own terms.
Beyond that separation, I look at whether campaigns are segmented in a way that maps to the business. A store with products across different margin tiers benefits from keeping those groups in separate campaigns, so budget and bidding targets can be set appropriately for each. A blanket ROAS target applied to a mix of high-margin and low-margin products means either overpaying for the low-margin items or underbidding on the high-margin ones.
I also check for campaigns competing with each other. It is common to find a Standard Shopping campaign and a Performance Max campaign running simultaneously with overlapping product coverage. They can bid against each other in the same auctions, raising costs for the same inventory.
The Third Layer: Search Terms and Negative Keywords
For accounts with Search campaigns, I pull the search terms report. This shows every actual query that triggered an ad, not just the keywords being targeted.
What I typically find: a meaningful percentage of spend going to queries unrelated to the store’s products, because match types were set broadly early on and never tightened, and negative keyword lists were never built up over time.
I have reviewed accounts with two or three years of history and fewer than 50 negative keywords. A well-managed account with that history typically has several hundred, built incrementally from regular search term reviews. The difference in wasted spend between those two situations is significant and direct. Every dollar captured by an irrelevant query is a dollar not available for a query that could convert.
The Fourth Layer: Bidding and Budget Allocation
I look at how budget is distributed across campaigns, whether that distribution matches the business priority of each campaign, and whether the bidding strategies in use are appropriate for the data available.
Smart bidding strategies like Target ROAS require a minimum level of conversion data to function well. Below roughly 30 conversions per month in a campaign, the algorithm does not have enough signal to make reliable predictions. Accounts that use Target ROAS with thin conversion data frequently oscillate, spending heavily when they should hold back, then cutting spend when an opportunity is there.
I also look for budget imbalances: campaigns that consistently exhaust their daily budget by early afternoon (a signal of either too-low budget or too-broad targeting), and campaigns with large budgets that perpetually under-spend because targeting is too narrow or bids are set too conservatively.
The Fifth Layer: Audience Signals and Data Quality
For Performance Max specifically, I look at what audience signals have been provided. Signals are the information you give Google to help it start finding the right customers. Most accounts I audit either have no signals configured (meaning Google starts from a very broad base) or have generic in-market audiences that don’t reflect actual buyer behavior.
What strong signals look like: a customer list from your email database or CRM, website visitors segmented by behavior (people who viewed a product, people who reached checkout, people who purchased), and any relevant demographic or interest profile that matches your actual buyer.
Accounts with better signal quality reach their target customers faster and waste less budget in the early phase of a new campaign. Accounts with no signals are teaching the algorithm from scratch, at your expense.
What Audits Usually Find
After reviewing a significant number of ecommerce accounts, certain patterns repeat.
Most accounts have at least two or three of the following: conversion tracking that is partially broken or measuring the wrong events, Performance Max running without brand exclusions so branded traffic inflates account ROAS, thin or absent negative keyword lists, bidding strategies mismatched to available data volume, no separation between new-customer and returning-customer performance.
The specific combination varies. But an account where everything is working correctly is uncommon. Problems accumulate quietly, especially in accounts that were set up by one person, handed off to another, and never reviewed end-to-end. Performance may be “good enough” that no one looks closely. But good enough is rarely optimal, and optimal is what you’re paying for.
Getting This Fixed
An audit is exactly the kind of engagement I do for ecommerce stores, either as a standalone review or as the starting point before any ongoing management. The findings typically identify recoverable budget and performance improvements that far exceed the cost of the review.
If you want to know what is actually happening in your account, I can help. Reach out at adnanagic.com/#contact.
Part of the Google Ads for Store Owners series, written for ecommerce owners who want their ad spend to actually work.
Related Posts
- How to Audit a Google Ads Account That Isn’t Performing
- What a Proper Ecommerce Google Ads Account Looks Like in 2026
- The Real Cost of Bad Conversion Tracking
Related Posts
What a Proper Ecommerce Google Ads Account Looks Like in 2026 vs What Most Stores Have
11 min read
Why Your Best-Selling Products Are Invisible in Google Shopping
10 min read
New Store, No Sales Data: How to Start Google Ads Without Burning Cash
10 min read
Need Help With Your Google Ads?
I help e-commerce brands scale profitably with data-driven PPC strategies.
Get In Touch