Use this checklist to audit any eCommerce Google Ads account. It covers the areas that most commonly limit performance — and the specific things to look for in each one. Work through it in order: tracking before structure, structure before bidding.
1. Conversion tracking
- Purchase conversion action is set as the primary conversion — not add-to-cart or checkout
- No duplicate conversion actions (native tag + GA4 import both recording purchases)
- Revenue value is populating dynamically with the actual order total
- Currency code matches your Google Ads account currency
- Confirmation page tag only fires after a genuine transaction (not on direct URL access)
- GA4 purchase event is passing transaction_id for deduplication
- Google Ads attributed revenue within 20–30% of actual platform revenue (discrepancy beyond that warrants investigation)
2. Account structure
- Branded search terms have a dedicated campaign (or are excluded from all non-brand campaigns)
- PMax campaign has brand exclusions applied
- No campaign is mixing brand and non-brand intent in the same budget pool
- Campaign count is appropriate to conversion volume (consolidation often beats fragmentation)
- Retargeting is separated from new customer acquisition campaigns
- Budget allocation reflects priority — not historical default
3. Performance Max
- Brand exclusions applied
- At least two asset groups — not one catch-all group for the whole catalogue
- Asset groups are segmented by product category or margin tier
- First-party audience signals applied (purchasers, cart abandoners, high-intent visitors)
- Each asset group has specific creative (not the same headlines/images everywhere)
- All asset types provided: headlines (15), descriptions (5), images (landscape, square, portrait), logo
- Video provided (not relying on auto-generated video)
- ROAS target based on contribution margin, not a round number or benchmark
4. Google Shopping / Feed
- Product titles lead with the most relevant search attributes (not internal product codes)
- GTINs populated for all branded products
- Colour, size, material attributes populated where applicable
- Item group IDs used to group variants
- Images are high resolution, product clearly visible, minimal background clutter
- Custom labels used for segmentation (at minimum: margin tier)
- Merchant Center disapproval rate below 5%
- Price and availability updating at least daily
- Out-of-stock products excluded from campaigns
5. Bidding strategy
- ROAS target is grounded in margin economics (not a wish or industry benchmark)
- Smart bidding campaigns have sufficient conversion volume (30+ conversions/month minimum)
- ROAS targets haven't been changed more than once in the past 4 weeks
- Budget is not capped below the point where it's limiting learning
- Bid adjustments (where applicable) are intentional — not legacy settings from years ago
6. Commercial performance
- New customer revenue is tracked separately from total revenue
- ROAS is being assessed by product tier, not just overall account average
- Low-margin products are either excluded or have higher ROAS targets applied
- Customer acquisition cost is compared against estimated customer lifetime value
- Brand vs. non-brand performance is reported separately
- Month-over-month and year-over-year trends are reviewed alongside absolute numbers
7. Measurement hygiene
- GA4 is linked to Google Ads (and data sharing is enabled)
- Audiences are being shared from GA4 to Google Ads
- Attribution model is understood and intentional (Data-Driven vs Last Click)
- Conversion windows match realistic purchase decision timelines
- A regular tracking verification check is scheduled (monthly minimum)
If you've worked through this list and found issues in three or more areas, you're likely dealing with a systemic account quality problem — not individual campaign issues. The fixes need to happen in order: tracking first, structure second, bidding last.