If your Shopping campaigns are spending steadily but revenue feels unpredictable, the problem is rarely just bids. More often, the issue sits deeper in the account: weak feed signals, poor product segmentation, wasted search terms, or a structure that gives Google too much freedom with too little control. That is exactly why knowing how to audit Google Shopping campaigns matters. A proper audit shows where spend is leaking, where margin is being ignored, and where scale is being blocked.
What a good Shopping audit is actually trying to find
A weak audit looks at ROAS, clicks and CPCs, then stops there. A useful one goes further. It asks whether the account is built to push profitable products harder, suppress poor traffic faster, and give your catalogue the best chance of matching to the right searches.
Google Shopping is not a standard keyword-led campaign type. You are trusting Google to interpret feed data, historical performance and user intent, then decide when your products should show. That means an audit has to assess both account management and product data quality. If either side is off, performance suffers.
For established eCommerce brands, that matters because spend inefficiency compounds quickly. A few percentage points of wasted budget in a large catalogue can become a serious drag on growth.
How to audit Google Shopping campaigns without missing the real problems
Start with the commercial context before you touch the account. If you do not know your margin profile, breakeven cost of sale, highest-value categories and stock priorities, the audit will be shallow from the start. A campaign can look healthy on platform metrics while still being poor for the business.
The first question is simple: are the campaigns aligned with business economics? If low-margin products are getting the same bidding aggression as high-margin products, or if clearance stock is buried while hero lines are constrained, the account is not being run commercially.
After that, work through the account in this order: feed health, campaign structure, search query quality, product segmentation, bidding logic, device and audience signals, then measurement. This sequence matters. There is no point criticising bidding strategy if the feed is weak or the campaign structure is fighting the catalogue.
Check feed quality before campaign settings
Shopping performance starts in the feed. Titles, product types, GTINs, descriptions, image quality, availability, pricing consistency and custom labels all influence how effectively products enter auctions and match to intent.
Look closely at titles first. Generic titles usually produce generic traffic. A title like “Men’s Trainers” is far less useful than one that includes brand, model, colour, size range or material where relevant. Better titles improve match quality. They also help separate high-intent searches from broad, low-converting traffic.
Then review product categorisation. If product types are vague, inconsistent or absent, you lose control over segmentation and reporting. Custom labels are equally important. Serious Shopping accounts use them to group products by margin, seasonality, top sellers, clearance status or price band. Without that layer, optimisation becomes blunt.
Also check for disapproved items, limited performance warnings and price mismatch issues in Merchant Centre. These do not just reduce volume. They distort the data you use to judge the account.
Audit the campaign structure for control, not neatness
A tidy account is not always an effective one. The real question is whether the structure gives enough control over budget, bidding and search intent.
If all products sit in one campaign with little segmentation, there is a high chance strong products are subsidising weak ones. That makes optimisation slower and hides waste. On the other hand, over-segmentation can create thin data, unstable bidding and unnecessary complexity. The right structure depends on catalogue size, budget and product economics.
In most cases, product grouping should reflect commercial priorities. Best sellers, high-margin ranges, seasonal products and strategically important categories should not be buried inside broad catch-all campaigns. They need room for distinct bidding and budget decisions.
You should also assess whether Standard Shopping and Performance Max are competing awkwardly. If both are running, the audit needs to examine overlap, reporting clarity and whether budget is being diverted into low-transparency inventory.
Search terms often reveal the biggest waste
One of the fastest ways to improve Shopping profitability is to review search term quality properly. This is where poor intent usually shows up first.
Look beyond surface metrics. A query may generate conversions but still be a weak fit if it drives low average order value, poor repeat purchase behaviour or a margin mix that does not work. Equally, some non-brand terms may look expensive at first glance but deserve patience because they support profitable new customer acquisition.
Focus on patterns. Are products showing for research-heavy queries with little buying intent? Are irrelevant modifiers appearing repeatedly? Is branded traffic inflating performance and masking weak non-brand activity? Are generic terms consuming budget that should be going to high-intent category or product-specific searches?
A disciplined negative keyword strategy still matters in Shopping. If the account has very few negatives, that is usually a warning sign. Google will happily spend into ambiguity unless someone stops it.
Product-level performance is where profit gaps appear
Account-level ROAS can hide a lot. A Shopping audit should always go down to product and category level.
Start by identifying the products spending heavily without converting at an acceptable cost. Then look at products converting efficiently but capped by low impression share or underfunded campaigns. These two groups usually tell you where the account is misallocating budget.
You also need to compare performance against margin, not just revenue. A product with a 600 per cent ROAS may still be mediocre if margins are slim. Another with a lower ROAS may be highly attractive if it supports stronger contribution profit or repeat purchase value.
This is where custom labels become commercially valuable. If you cannot quickly separate products by margin tier, stock priority or strategic role, optimisation turns reactive.
Review bidding strategy with a healthy level of scepticism
Smart Bidding can work well in Shopping, but it is not self-correcting magic. When auditing, examine whether the current bidding strategy actually fits the account’s data quality and volume.
If tROAS is set unrealistically high, volume may be throttled. If it is too soft, the campaign may expand into weak traffic. If Maximise Conversion Value is running without sensible guardrails, spend can drift towards easy but less profitable revenue.
Look for volatility as well. Sudden swings in spend, conversion value or impression share often point to unstable inputs, whether that is feed disruption, budget limitation, promotional pricing changes or poor campaign segmentation.
A common mistake is blaming bidding for what is really a structural issue. If query quality is poor or products are grouped badly, changing bid strategy will not solve the root problem.
Do not ignore audience, device and geography signals
Shopping is feed-led, but context still matters. During an audit, review whether audience data is being used to guide decisions rather than sit passively in observation mode.
Returning visitors, past purchasers and high-value customer segments often behave very differently from cold traffic. If the account treats them the same, there may be missed opportunities in both bidding and exclusions.
Device data deserves scrutiny too. Mobile may dominate clicks while desktop carries stronger conversion efficiency, especially in higher-ticket categories. That does not mean mobile is bad. It means you need to understand its role in the path to purchase.
Geography can also expose waste. Some regions convert profitably; others drain budget through high delivery costs, weak conversion rate or poor stock fulfilment economics. If location-level performance has not been reviewed in months, the audit is incomplete.
Measurement problems can make a decent account look good or bad
Before making big decisions, verify the tracking. Shopping audits fall apart when conversion data is inflated, duplicated or stripped of value accuracy.
Check whether purchases are being recorded once, whether revenue matches the platform closely enough, and whether refunds, VAT treatment and shipping values are handled consistently. If offline margins differ materially by product group, platform revenue alone is not enough for decision-making.
For more mature brands, the best audits bring in backend data. That includes contribution margin, new versus returning customer performance, and category-level profitability. Google Ads should inform decisions, not define them in isolation.
What usually needs fixing after the audit
In most accounts, the same themes keep coming up. Feed data is too weak to support strong matching. Campaigns are segmented by convenience instead of margin logic. Search term control is light. Budget is being spread too evenly. Strong products are not isolated enough, and weak ones are allowed to spend for too long.
The fix is not usually dramatic. It is disciplined. Better titles. Smarter custom labels. Cleaner exclusions. More commercial segmentation. Bidding targets grounded in actual economics, not wishful ROAS numbers.
That is the difference between an account that merely spends and one that scales profitably. Oxedent’s view is simple: Shopping campaigns should be judged by profit contribution and growth quality, not by whether the dashboard looks busy.
A good audit should leave you with a sharper question than “How do we get more sales?” It should tell you which products deserve more spend, which queries need shutting down, and where the account is making growth harder than it needs to be.