PerformanceMaxforShopify:TheCommercialPlaybook

Why most Shopify brands lose money on PMax

Performance Max is a black-box campaign type. Google decides which audience, placement, creative and bidding strategy to deploy from the assets you give it. On a clean account with margin-aware inputs, that delegation produces strong results. On a typical Shopify account it produces revenue without profit.

The structural problem: PMax optimises for the conversion value you feed it. If the conversion value is gross order value, the algorithm bids hardest on the SKUs with highest revenue per click — which on most catalogues are not the same SKUs as highest margin per click. Best-sellers on Shopify are often loss-leaders, sale items or low-margin staples. PMax piles spend into them and the P&L deteriorates while ROAS looks fine.

The Shopify-PMax integration stack

The technical setup that makes PMax viable on Shopify, in dependency order:

1. Google & YouTube channel app installed in Shopify, with the product feed syncing to Google Merchant Center. This is the baseline; without it there is no Shopping inventory for PMax to surface.
2. Cost-per-item populated at SKU level in Shopify (Products → Inventory → Cost per item). This is the field that becomes the margin signal everything downstream relies on.
3. Custom labels in the feed for commercial intent: SKU job (Scale / Profit / Protect / Recovery / Gateway), margin band (high / medium / low) and inventory health (in-stock / low / overstock).
4. Enhanced conversions for web turned on in Google Ads with first-party data passed through. This restores attribution lost to consent mode and iOS privacy changes.
5. Conversion value rules that adjust order value by margin band — so the algorithm bids on profit, not revenue.
6. Brand exclusion list applied at account level. Non-negotiable; PMax will eat branded search otherwise.

Account structure that survives a P&L review

The default agency playbook is one PMax per product category. That structure fails for Shopify because product categories rarely correspond to commercial intent. A category like "candles" can contain Scale SKUs, Profit SKUs, Recovery SKUs and Gateway SKUs simultaneously — the algorithm cannot bid differently across them when they live in the same campaign.

The structure we use across JudeLuxe Shopify accounts:

• Branded Search campaign: exact, phrase and BMM match types for the brand name and obvious variants. High bid, high impression share target. Sits outside PMax.
• Scale PMax: high-margin growth SKUs. Target ROAS set to deliver the commercial POAS target for the Scale job — typically 1.5–2.0x POAS depending on category.
• Profit PMax: margin-protection SKUs. Conservative target ROAS, sometimes maximise conversion value with a strict cap.
• Gateway PMax: acquisition SKUs that bring new customers in. LTV-aware target — first-order POAS may be sub-1.0x deliberately, justified by 90-day cohort economics.
• Recovery PMax (only when relevant): SKUs being de-prioritised. Tightly capped, sometimes paused entirely. Used to drain inventory without scaling spend.

Product category splits live inside each PMax as asset groups, not as separate campaigns. This is the structural difference between agencies that treat PMax as a Shopping campaign and agencies that treat it as a commercial allocation tool.

The minimum spend conversation nobody wants to have

PMax has a structural floor. Each asset group needs roughly 30+ conversions per 30 days to exit learning. A campaign with three asset groups therefore needs ~90 conversions per month. At a £40 AOV that is £3,600 of revenue per group at a 100% close rate, which never happens — realistically £12,000–£20,000 of revenue per group, meaning £30,000–£60,000 of revenue across the campaign.

Run the maths backwards from your conversion rate and AOV. If the spend required to clear that volume threshold exceeds your monthly budget, PMax is the wrong campaign type for the account — Standard Shopping or Search campaigns will deliver more stable results below the threshold. The brands quietly losing money on PMax below £4,000/month are almost always running an under-resourced campaign type, not a broken one.

JudeLuxe's published minimum is £10,000/month in ad spend before we recommend PMax as the primary campaign type. That figure is not arbitrary — it is the threshold below which the bidder's behaviour becomes too volatile to manage commercially.

What to do in week one

1. Pull a search terms report (via Insights → Asset groups → Search themes) and document everything PMax served against. If branded terms are in there, that is your first leak.
2. Audit the cost-per-item field in Shopify. Spot-check 20 SKUs. If costs are missing, blended margin signals are unreliable.
3. Build a SKU job table in a spreadsheet. Assign each top-50-revenue SKU one of Scale / Profit / Protect / Recovery / Gateway. This is the input for custom labels.
4. Apply the brand exclusion list at account level. Submit the request to your Google rep if you do not have direct access — it is a one-form process.
5. Set conversion value rules for the next 30 days. Even a coarse three-band rule (high-margin +20%, medium 0%, low −30%) reshapes bidder behaviour materially.

Where Shopify makes this easier than Magento or WooCommerce

Shopify's product schema is well-formed and the Google & YouTube channel app handles most of the feed-syncing complexity. Magento brands typically need a third-party feed tool (Channable, GoDataFeed) and WooCommerce brands face plug-in fragmentation. For Shopify, the feed work is largely solved out of the box — which is why the remaining problems are commercial rather than technical.