A profitable account can look inefficient at campaign level. An unprofitable account can look healthy in a platform report. That is why you do not audit ecommerce PPC account performance by scanning ROAS columns and declaring a winner. You audit it by tracing every pound of spend back to product margin, customer value, tracking quality and the commercial objective behind the campaign.
For established retailers, the point of an audit is not to create a longer task list. It is to identify where budget is being wasted, where profitable demand is being capped and what must change before you scale spend.
Start with the commercial numbers, not the ad account
Before looking at Google Ads, Meta or Performance Max, establish what success actually means. You need a clear breakeven cost of sale or target ROAS by product range. A blended target for the whole business is useful, but it is rarely enough on its own.
A £30 product with a tight contribution margin cannot be judged by the same acquisition target as a £150 bundle with repeat purchase potential. Returns, shipping costs, payment fees, discounts and VAT can all change the picture. If these inputs are vague, any optimisation decision becomes guesswork dressed up as reporting.
Review revenue by channel alongside blended MER, contribution margin and new versus returning customer share. Platform-attributed revenue has a role, but it is not the business result. If paid revenue rises while total profitable growth does not, attribution may be over-crediting ads that were already capturing demand.
Verify tracking before judging performance
Poor tracking is one of the most expensive problems in ecommerce PPC because it causes good campaigns to be cut and weak campaigns to be funded. Check that purchase values, currencies, transaction IDs and product-level data pass correctly from the site into each advertising platform.
For Google Ads, inspect conversion actions, attribution settings, enhanced conversions and whether duplicate purchases are being recorded. For Meta, review Pixel and Conversions API event matching, event prioritisation and purchase value consistency. A purchase event that fires twice, reports tax inconsistently or loses value through a checkout issue will corrupt optimisation.
Do not stop at the platform setup. Compare a sensible date range of platform revenue with your ecommerce platform and analytics data. The figures will not match perfectly. Different attribution windows and consent choices guarantee that. What matters is whether the gap is understood, stable and small enough to make confident budget decisions.
Check the signals your campaigns are optimising towards
Automated bidding can only act on the signal it receives. If a Performance Max campaign is optimising for every low-value sale equally, it will chase volume rather than profit. If Meta is fed only purchase events without customer or margin context, it may find cheap customers who never buy again.
Where the business has the data maturity, bring more commercial intelligence into the measurement plan. That could mean separating new and returning customer performance, excluding cancelled orders or using profit-adjusted values for key product groups. It is not necessary to over-engineer an account, but it is necessary to stop treating every pound of revenue as identical.
Audit ecommerce PPC account structure for control
Campaign structure should give you enough control to make commercial decisions without fragmenting data so badly that automation cannot learn. This balance matters more than any fashionable account template.
In Google Ads, examine whether Search, Shopping and Performance Max have clearly defined roles. Brand campaigns should not be allowed to hide weak non-brand acquisition. Shopping and Performance Max should be assessed by product profitability, not merely their combined headline ROAS. Search campaigns need a search term review that shows whether spend is buying high-intent demand or broad, expensive research queries.
A common issue is a legacy structure built around old platform recommendations rather than the current catalogue and margin profile. Hundreds of thin ad groups, overlapping campaigns and duplicated product targets create noise, not sophistication. Equally, collapsing everything into one campaign can conceal the products that deserve more investment and those that should be restricted.
On Meta, assess whether prospecting, retargeting and creative testing can be read clearly. Excessive audience splits often prevent delivery, while one catch-all campaign can make it impossible to understand what is driving results. The right answer depends on spend, catalogue size, creative output and conversion volume. The principle stays the same: structure must support profitable action.
Follow the money through the product feed
For retailers using Shopping or Performance Max, the product feed is not an admin task. It is a performance asset. Weak titles, generic images, missing attributes and inaccurate availability reduce relevance before bidding even begins.
Audit titles against the way customers search. The strongest titles usually lead with the product type and include meaningful differentiators such as brand, size, material, colour or compatibility where relevant. Avoid stuffing every possible keyword into a title. The goal is accurate matching, not unreadable product data.
Then segment performance by product ID, category, brand and price point. Look for products with strong conversion rates but limited impressions, products consuming spend without sales, and high-return lines that inflate apparent revenue. Bestsellers are not automatically the best products to advertise. A low-margin bestseller can drain budget that a higher-margin, lower-volume product could use more profitably.
Also check feed disapprovals and warnings. A small number of unavailable, mispriced or policy-limited products can remove valuable inventory from auctions without being obvious in a headline report.
Find wasted spend without cutting growth
Waste reduction is not the same as pausing everything below target ROAS. Some campaigns assist a sale, introduce new customers or need more data before a reliable decision can be made. But recurring waste is usually visible when you look beyond the top line.
Review search terms, placement reports, product spend, geo performance, device trends and time-of-day patterns where volume supports it. Pay particular attention to spend that produces clicks but no meaningful progression through the funnel. This could indicate irrelevant queries, poor landing-page alignment, weak pricing or a site experience problem rather than an advertising issue.
Use exclusions carefully. Broadly excluding placements or search themes because they had one bad week is not disciplined management. Decisions should reflect enough spend and conversion data to be credible. The opposite mistake is allowing obviously irrelevant traffic to run for months because nobody has ownership of the account.
Review bidding, budgets and lost opportunity
An account can miss its target because it is spending too much. It can also miss it because profitable campaigns are constrained while budget sits in weaker activity. Audit budget allocation as aggressively as you audit waste.
Check whether high-performing campaigns are limited by budget, product availability or bid targets that are too restrictive. A target ROAS set unrealistically high can suppress volume and leave profitable revenue on the table. Lowering the target may reduce reported ROAS while increasing total contribution. That is a trade-off worth making when the numbers support it.
Conversely, raising budgets too quickly can destabilise delivery, especially when conversion volume is modest. Scale in measured increments, watch marginal performance and protect the campaigns that have earned more investment. There is no universal percentage increase that works for every account.
Inspect creative and landing-page alignment
PPC performance is not created inside the ad platform alone. If ad creative promises a clear benefit and the landing page buries it beneath slow load times, unclear delivery information or a weak product page, the account will pay more for every conversion.
For Meta, look at creative fatigue, message variation and whether ads give a customer a reason to choose the product now. For Search and Shopping, assess whether the landing page answers the query, reflects the advertised price and presents the right variant without friction. Creative should be tested against commercial outcomes, not just click-through rate. A high-click ad that attracts bargain hunters can be expensive theatre.
Turn findings into a prioritised plan
A useful PPC audit ends with a small number of actions ranked by financial impact, confidence and effort. Fix broken tracking before rebuilding campaigns. Resolve feed errors before debating minor bid adjustments. Move budget from proven waste into proven opportunity before launching speculative tests.
Document the baseline, the expected impact and the metric that will confirm whether the change worked. This protects the business from constant account tinkering and gives everyone a clear view of accountability. It also makes agency management easier to judge: you should be able to see what changed, why it changed and what commercial result followed.
The best time to audit is before performance becomes a crisis. If your ecommerce PPC account has spend, product-market fit and a clear growth target, an honest audit can turn scattered activity into a profit-led plan – and give every additional pound of budget a job to do.