Google Ads Audit Checklist for Ecommerce

What should you collect before opening the account?

Set the audit window before judging performance. Use a recent period long enough to contain normal demand, compare it with the prior period, and add the same period last year when seasonality matters. Export store orders, refunds, revenue, gross margin by product, stock status and any promotion calendar for the same dates.

Then write down the account's actual job. Is it acquiring new customers, harvesting branded demand, clearing inventory, or maximizing contribution profit? A campaign can hit its platform target and still work against the business goal. The audit needs a declared outcome and a source of truth outside Google Ads.

For a structured outside review, Barlo Digital's ecommerce ads audit checks 30 days of Google and Meta spend and returns a prioritized action list.

1. Can you trust the purchase and revenue data?

Start by placing a test order. Confirm that the purchase fires once, carries the correct currency, order value and transaction ID, and reaches the conversion action used for bidding. Reconcile Google Ads purchase value against the store's order revenue for the same dates. Document known differences such as attribution windows, refunds, tax and shipping instead of forcing an exact match.

In Goals, inspect every conversion action. Purchases should not compete with page views, add-to-cart events or other micro-actions as equal bidding goals. Check action source, counting method, value setting, attribution model and whether enhanced conversions are configured. Do not change bidding until material discrepancies are understood.

Bad looks like duplicate purchases, static values, missing transaction IDs, several tags recording the same sale, or a campaign optimizing toward an event that is not revenue. Treat those as blockers, not ordinary optimization ideas.

2. Do targets reflect margin and customer economics?

Translate the business constraint into a break-even target. A revenue-based target needs gross margin, expected returns, payment fees, fulfillment costs and any acceptable acquisition investment. A single account-wide return on ad spend target can misallocate budget when margins differ sharply by product.

Example: if an order produces $40 of contribution before advertising, spending more than $40 to acquire that order loses money on the first purchase. If repeat purchases are part of the plan, state the payback window and use a defensible customer value rather than an assumed lifetime value.

External averages are context, not targets. A 2026 dataset drawn from 25 ecommerce Google Ads accounts reported Performance Max averages of $1.53 CPC, 3.32% conversion rate, $46.10 CPA and 4.37x ROAS. The source also says the accounts spanned four countries and that the number of active accounts changed over time, which limits direct comparison. Review the dataset and methodology.

3. Does each campaign have one clear job?

Map every active campaign to demand type, inventory and business objective. Separate branded demand in reporting so it cannot hide weak customer acquisition. Look for the same products or queries competing across Search, Shopping and Performance Max without a deliberate reason.

For Search, inspect match types, location settings, network settings, audiences, ad schedules, final URLs and bid strategy. For Shopping and Performance Max, inspect listing groups, asset groups, product exclusions, brand controls and campaign priority logic. A clean account does not require many campaigns. It requires explicit roles and minimal accidental overlap.

Google says Performance Max can serve across Search, Maps, YouTube, Discover, Gmail and the Google Display Network. Its channel report can show clicks, conversions and cost by channel, so use that view before treating the campaign as one opaque total. See Google's Performance Max evaluation guidance.

4. Which search terms are consuming money without intent?

Sort search terms by cost, then label each as relevant non-brand, brand, competitor, informational or irrelevant. Review both individual terms and themes. Add negatives only when the intent is clearly wrong. A broad exclusion can remove useful demand along with the waste.

Performance Max now exposes granular search-term reporting. Check the term, landing page and ad format together. Google notes that historical Performance Max search-term data begins in March 2023, so do not interpret an earlier blank period as zero activity. Read Google's search-terms report documentation.

Bad looks like brand traffic reported as new-customer acquisition, irrelevant queries with repeat spend, search themes that add no useful reach, or landing pages that do not satisfy the query.

5. Is the product feed limiting eligible demand?

Open Merchant Center diagnostics and separate disapprovals from warnings and optimization opportunities. Check title accuracy, description, GTIN, product category, item group ID, color, size, image, price, availability, shipping and returns. Then segment Google Ads performance by product ID. Identify profitable products with little exposure and high-spend products with weak contribution.

Google states that missing or incorrect product information can cause disapprovals, limited eligibility or incorrect displays. Its examples include GTIN, item group ID, color, size, images and conflicts between the feed and website. Use Google's product data specification as the validation reference.

A valid feed is only the starting point. Product titles should communicate the attributes shoppers use, and campaign groupings should reflect margin, stock and product role rather than arbitrary catalog categories.

6. Is budget pacing aligned with profitable opportunity?

Compare daily spend with plan, but do not judge pacing alone. Pair spend with marginal return, inventory and conversion lag. Find campaigns that regularly hit budget while meeting the business target, campaigns that spend freely below target, and profitable products blocked inside a campaign with weaker inventory.

Use the budget report, bid-strategy status and change history. Check whether target return or target CPA changes, budget edits, seasonality adjustments, experiments or automated recommendations explain a performance break. A sudden result is not evidence of an algorithm problem until the timeline rules out account changes, tracking changes and demand shifts.

7. Do ads and landing pages complete the same promise?

For each material query or product group, read the ad and destination as one experience. Confirm that the landing page shows the expected product, price, availability, shipping information, returns policy and a usable mobile checkout path. Review final URL expansion and automatically created assets so traffic does not land on an unintended page.

Compare landing pages by cost, conversion value and conversion rate, but investigate mix before declaring a page weak. Brand traffic, devices, countries and product prices can create large differences. The useful finding names the segment, the page problem and the money exposed.

How should audit findings be ranked by dollars?

Turn each observation into a decision record. Include the evidence, affected spend or revenue, confidence, effort, owner and next measurement date. Estimate impact as a range when attribution is uncertain. Never present the estimate as a guaranteed result.

Example: if an irrelevant search-term group spent $600 in the review period, the documented exposure is $600. The savings are not automatically $600 because exclusions can change traffic and conversion mix. Record the spend, make the narrowest safe change, and measure the next comparable period.

What belongs in the final PPC audit checklist?

• Test purchase completed and recorded once with correct value and transaction ID.
• Store orders and revenue reconciled with platform reporting.
• Primary conversion goals limited to meaningful commercial outcomes.
• Return and acquisition targets tied to margin and payback assumptions.
• Brand, non-brand and new-customer performance visible separately.
• Campaign roles, overlap, networks, locations and final URLs reviewed.
• Performance Max channels, search terms and product results inspected.
• Merchant Center disapprovals and key product attributes checked.
• Budget constraints and bid-strategy status reviewed with change history.
• Ads, assets and landing pages checked as one customer path.
• Every finding ranked by evidence, dollar exposure, confidence and effort.