The Profit-First Google Ads Account Structure for Ecommerce in 2026

Revenue-first structures are a 2018 idea

If you opened your Google Ads account today and saw one giant Performance Max campaign with every product in one feed, optimizing for “Maximize Conversion Value,” you are running a 2018 setup in a 2026 market.

It might still be making money. It is almost certainly not making the money it could.

The brands scaling profitably in 2026 have rebuilt their Google Ads accounts around a single question: which products, at which margins, deserve which level of spend? Everything else - campaign types, bidding strategies, feeds - follows from that.

The profit-first principle

Treat your product catalog like a portfolio. Some products are loss leaders. Some are margin engines. Some are seasonal. Some are brand defenders. Google has no idea which is which unless you tell it.

The profit-first structure groups products by margin profile, sets bidding targets that reflect actual profit, and uses campaign architecture to prevent the algorithm from spending your most flexible budget on your weakest products.

The five-campaign skeleton

For a typical Shopify or Woo store with €30k-€500k monthly spend, the structure looks like this:

1. Brand Search. Defensive. Captures branded queries cheaply. Always exact match, separated from generic search.
2. Generic Search. Non-brand, high-intent keywords. Manual structure with tight match types and negative lists.
3. Performance Max - Hero SKUs. Your top 10-20% of products by profit contribution. Separate feed, separate budget, aggressive tROAS.
4. Performance Max - Long Tail. Everything else, lower priority, defensive budget. Catches incidental demand without cannibalizing the hero campaign.
5. Shopping or PMax - New Customer Acquisition. Optional but powerful. Uses Google’s new-customer-only bidding to pay more for first-time buyers.

If you are running international, replicate this skeleton per market. Do not combine countries in one campaign - the bidding math breaks.

Why splitting hero SKUs from long tail matters

If you put your best margin products in the same campaign as the rest of your catalog, Performance Max will spend most of the budget on whichever products it can sell the most of, not on whichever products make you the most profit. Often these are not the same products.

Splitting forces a budget decision. You decide how much goes behind the hero SKUs versus everything else. The algorithm decides how to spend it within each bucket. Both you and the algorithm are doing what you are good at.

Bidding to profit, not revenue

Default tROAS targets are calculated on revenue. That is fine if your margin is uniform. It is dangerous if it is not.

The fix is to feed Google your profit data, not just your revenue data. Two approaches work in 2026:

• Conversion value adjustments. Send Google an adjusted conversion value equal to gross profit instead of order value. The algorithm now optimizes for profit directly.
• Profit-aware feeds. Use tools like ProfitMetrics, Adsmurai, or a custom feed app to inject a custom_label for profit tier per product, then bid differently per tier.

Either way, the input changes from “how much revenue did this click drive?” to “how much profit did this click drive?” The bidding decisions get dramatically better.

Custom labels: the underused lever

custom_label_0 through custom_label_4 in your product feed are the most underused fields in Google Ads. Use them to tag:

• Margin tier (high / medium / low)
• Inventory level (in stock / low stock / overstocked)
• Lifecycle stage (new / hero / clearance)
• Seasonality (evergreen / seasonal)
• Bestseller flag

Each label becomes a filter you can use to build campaigns. Want a PMax campaign that only promotes overstocked items at higher tROAS? Done. Want to exclude clearance items from new-customer campaigns? Done.

Search and brand still earn their keep

Even with Performance Max dominating ecommerce conversations, plain Search campaigns still matter. Brand Search is the cheapest revenue you will ever buy. Non-brand Search lets you compete on high-intent terms with full transparency, something PMax intentionally hides.

Run them. Just structure them tightly:

• Separate campaigns for brand and non-brand.
• Negative keyword lists shared across the account.
• Exact and phrase match, no broad match without an experimental budget.
• Tight ad group themes - five to ten keywords each, not five hundred.

The reporting layer

A profit-first structure only works if you can see profit per campaign. That means your reporting stack needs:

• Server-side tracking sending true conversion values to Google.
• A profit calculation layer (ProfitMetrics, Triple Whale, custom Looker Studio).
• A weekly review where every campaign is judged on profit ROAS, not revenue ROAS.

Without this, you cannot tell whether the structure is working. With it, the budget decisions become obvious.

Migration plan

If your account is currently one big PMax campaign, do not blow it up overnight. Migrate in three phases:

1. Week 1-2. Set up the reporting layer. Calculate profit per product. Identify hero SKUs.
2. Week 3-4. Add Brand Search, Generic Search, and a new Hero PMax campaign in parallel. Let them learn while the original campaign keeps running.
3. Week 5-8. Shift budget gradually from the original campaign into the new structure. Decommission the old campaign once the new ones are stable.

Done well, the migration typically lifts profit by 15-30% within a quarter without increasing spend. The revenue number may stay flat or even dip. The profit number, which is the one that actually matters, goes up.

The principle to remember

Google’s algorithm is a powerful, blunt instrument. Your job is not to outsmart it. Your job is to point it at the right products with the right targets and the right data. A profit-first structure is how you do that.