Most Google Ads accounts for eCommerce brands accumulate structure rather than design it. A Shopping campaign gets added. Then a Search campaign. Then Performance Max at some point because Google pushed it. Then some retargeting. Then some Display. The result is a collection of campaigns that sort of work individually but don't form a coherent system.
When we build an account from scratch, we start from a set of questions that most account builds skip: what are we actually trying to achieve, what does the funnel look like, and how do we want to allocate budget across different customer intents? The structure follows from those answers.
The four layers of intent we design for
Every customer arriving at an eCommerce site through Google is somewhere on an intent spectrum — from just becoming aware of a category to ready to buy right now. A well-structured account maps to that spectrum rather than just chasing conversions indiscriminately.
- Brand intent: People searching for you by name. High conversion rate, low acquisition value (they already know you). Defended with a branded Search campaign.
- High purchase intent: Product searches, competitor searches, high-specificity category searches. Standard Shopping or PMax asset groups targeting proven query types.
- Research intent: Broad category queries, comparison queries, "best X for Y" searches. Usually lower converting but important for top-of-funnel awareness at scale.
- Retargeting: Past visitors, cart abandoners, past purchasers. Different messages, different bids, different expectations.
The structural implication is that these layers shouldn't compete with each other. Mixing brand and non-brand into one campaign, or blending retargeting with new customer acquisition, obscures performance and forces compromises in bidding that benefit neither.
The baseline structure we use
Branded Search campaign
A dedicated campaign targeting exact and phrase match brand keywords. This protects brand search traffic at controlled cost and prevents PMax from claiming credit for it. Target CPA or manual CPC, limited budget — brand traffic converts well and doesn't need aggressive spend.
Performance Max campaign (with brand exclusions)
For most established eCommerce brands, PMax is the primary non-brand acquisition channel. It has access to all Google inventory and uses feed data, audience signals, and creative assets to find buyers across the funnel. We always apply brand exclusions so it doesn't cannibalise the branded Search campaign.
Asset groups within PMax are segmented by product category and margin tier where the catalogue warrants it. Each asset group has tailored creative specific to the products it covers.
Standard Shopping campaigns (selective use)
In some accounts, particularly those with complex product catalogues or specific products that need precise bid control, Standard Shopping campaigns sit alongside PMax for those SKUs. This isn't always necessary — PMax handles Shopping well when fed properly — but it's a useful tool for high-value products where you want more transparency and control.
Search campaigns for high-value non-brand queries
For brands selling considered-purchase products (higher AOV, longer decision cycles), dedicated Search campaigns targeting high-intent non-brand queries often outperform PMax for those query types. The control over ad copy, landing page, and bid adjustments by query is valuable when each conversion is worth a lot and you want to be precise.
The feed is part of the structure
Account structure doesn't end at the campaign level. For Shopping and PMax, the product feed is a structural input — and a poorly structured feed undermines whatever campaign architecture you've built on top of it. Before launching, we audit:
- Title optimisation — are product titles leading with the most important attributes for search matching?
- Attribute completeness — are all required and recommended attributes populated?
- Custom labels — are products tagged with the segmentation logic we need (margin tier, category, seasonality)?
- Image quality — are images clean, high-resolution, on-brand, and showing the product clearly?
What good structure enables
A well-structured account is easier to diagnose when something goes wrong, easier to optimise systematically, and produces cleaner data for decision-making. The campaigns tell a story: this is our brand defence, this is our new customer acquisition, this is our retargeting. Each has a clear job. Performance issues can be traced to specific layers rather than drowning in aggregated campaign data.
The other benefit is that good structure compounds. A campaign designed with clear intent, proper exclusions, and the right data inputs performs better from day one and improves faster as the algorithm learns. The accumulated advantage over an account that was never properly designed is significant — and it shows up in the numbers over time.