How should I structure a Google Ads account for eCommerce?

Simpler than most accounts are. A clean baseline is three parts: a dedicated Brand Search campaign to capture your own name cheaply; Performance Max or Standard Shopping for the catalogue, driven by your product feed; and selective non-brand Search for a few high-intent generic terms. Underneath, set the conversion goal to Purchase with real revenue values. Structure should follow how you want to control budget and read results, not how many campaigns you can create.

Should I separate brand and non-brand campaigns?

Yes. Brand traffic (people searching your name) is cheap and converts extremely well, so it should sit in its own campaign where you can see and control it. Mixing brand into non-brand or letting Performance Max swallow it flatters your numbers and hides how your real prospecting is performing. Separating brand is one of the highest-value structural moves in an eCommerce account.

How many campaigns does an eCommerce account need?

Usually far fewer than people build. Smart bidding needs a steady flow of conversions per campaign to learn, so splitting a small account into many campaigns starves each one of data and makes results erratic. Most small and mid-sized stores run well on a handful of campaigns: brand, a catalogue campaign (PMax or Shopping), and one or two focused non-brand campaigns. Consolidate rather than fragment.

Should I use one campaign or ad group per product?

Almost never. A campaign per product (or the old single-keyword ad group approach) fragments your conversions across dozens of tiny campaigns, none of which gets enough data for smart bidding to work. Group products by margin, priority or product type instead, and let the feed and automation handle the product-level detail. Structure by how you want to steer budget, not by SKU.

Performance Max or Standard Shopping for eCommerce?

Both use your product feed; they trade reach for control. Performance Max runs across all of Google's surfaces and hands placement and bidding to automation, with less reporting. Standard Shopping stays in Shopping placements and gives you more visibility and control over search terms and structure. Many accounts use PMax for broad feed-driven reach and keep or add Standard Shopping where they want tighter control; the right mix depends on how much transparency you need.

How to Structure a Google Ads Account for eCommerce (2026), Watermouse

The short answer

Good structure is simpler than most accounts are. A clean eCommerce baseline is three parts: a dedicated Brand Search campaign, a feed-driven catalogue campaign (Performance Max or Standard Shopping), and selective non-brand Search for a few high-intent terms, all on a Purchase goal with real revenue values. Structure should follow how you want to control budget and read results, not how many campaigns you can create. Fewer, well-fed campaigns beat many fragmented ones.

Ask ten people how to structure a Google Ads account and you will get ten diagrams, most of them far too complicated. In 2026, with smart bidding doing the heavy lifting, structure has one job: give you control and a clean read while giving the automation enough data to learn. Everything below follows from that. Here is the structure I actually build for an eCommerce brand.

What structure is really for

Structure is not about micromanaging keywords any more. It exists to do three things:

Control, so you can move budget toward what works and away from what does not.
A clean read, so you can tell brand from prospecting and one product group from another.
Enough data per campaign, so smart bidding has the conversions it needs to learn.

Every over-complicated account fails the third one: it slices a modest number of conversions across so many campaigns that none of them can learn. Keep that tension in mind and most structure questions answer themselves.

The baseline eCommerce structure

For most small and mid-sized stores, this is the whole thing:

Layer	What it does
Brand Search	Captures people searching your name, cheap, high-converting, and protects you from competitors bidding on your brand.
Catalogue (PMax or Shopping)	The engine. Feed-driven, puts your products in front of buyers across Google. This is where most spend and revenue live.
Non-brand Search (selective)	A small number of high-intent generic terms where you want tight control that a feed campaign cannot give.

That is usually a handful of campaigns, not dozens. You add complexity only when a real need appears (a second market, a product line with very different economics), never for its own sake.

Always separate your brand traffic

This is the single highest-value structural move, and the most commonly skipped. People searching your brand name are cheap to reach and convert extremely well, they were already coming to buy. If that traffic is buried inside non-brand campaigns or swallowed by Performance Max, it inflates those campaigns' results and hides how your actual prospecting is doing.

Do this: give brand its own Search campaign so you can see it, control it, and stop it from flattering everything else. When you later judge whether a campaign is truly winning new customers, you will be glad brand is walled off on its own.

The catalogue: PMax or Standard Shopping

Your products reach buyers through your Shopping feed, run either by Performance Max or Standard Shopping. Both use the same feed; they trade reach for control. The key structural decision is not the campaign type so much as how you split the catalogue, and the answer is: as little as possible, and by economics, not by SKU. Sensible ways to split, only if you need to:

By margin or priority, so your high-margin or hero products can carry a different target than thin-margin lines.
By product type, when different ranges genuinely behave differently.

What to avoid is a campaign or ad group per product. It fragments conversions into dust and starves the bidding. Group broadly, feed it well, and let automation handle the item-level detail.

Non-brand Search, used sparingly

There is still a place for classic Search on non-brand terms, but a smaller one than there used to be. Reserve it for a few high-intent generic queries where you want explicit control over the keyword and the message, rather than trusting a feed campaign to find them. Keep the themes tight and the count low. If a non-brand Search campaign is not getting enough conversions to bid well, it is a candidate to fold back in, not to split further.

The rule that governs all of it: consolidate, do not fragment

Almost every structural mistake in an eCommerce account is the same mistake, too many campaigns. It feels organised, but it quietly breaks smart bidding by leaving each campaign with too few conversions to learn from. The instinct to add another campaign for every idea is exactly the one to resist.

The heuristic: before creating a new campaign, ask whether it will get enough conversions each week to bid well on its own. If not, it belongs inside an existing campaign. Fewer, well-fed campaigns almost always beat a tidy-looking sprawl of starved ones.

The part underneath the structure that actually matters

No structure saves a broken measurement setup. Before you admire your campaign tree, make sure the foundation is right: the conversion goal is Purchase (not add-to-cart or pageviews), it carries real revenue values so your ROAS target means something, and you have a clear source of truth (which is also why Google Ads and GA4 never quite agree). Good structure on top of bad tracking is a tidy house built on sand.

Not sure whether your account is cleanly structured or quietly fragmented, with brand inflating your numbers? A free audit maps your actual structure, shows where it is starving the bidding or hiding your brand, and what a simpler build would change.

How to structure a Google Ads account for eCommerce