Google's own help center has spent 2026 telling advertisers to simplify. The "ABCs of Account Structure" guidance is explicit: move single-keyword ad groups into themed groups, drop granular match types in favor of broad match, and stop segmenting by device the way accounts did five years ago, according to Google's current account structure guidance.
That advice is correct, as far as it goes. It is also being applied one layer too high by a lot of ecommerce accounts, and that is where Google Ads account structure stops being a housekeeping question and starts quietly breaking the reporting a founder makes budget decisions on.
An account has three layers, and "simplify" is the right instinct for exactly one of them.
- Keywords and ad groups: consolidate. Google's own guidance and Smart Bidding both work better with fewer, broader, well-themed groups than with the old one-keyword-per-ad-group model.
- Campaigns: segment, but only where two things genuinely need their own budget, their own bidding goal, or their own line in a report. Collapsing campaigns for tidiness is where ROAS numbers stop meaning anything specific.
- Naming and conversion actions: this is pure measurement plumbing. Get it wrong and every layer above it reports a number that cannot be reconciled against Shopify.
Where "simplify everything" goes wrong
Smart Bidding optimizes toward whatever a campaign's budget and goal tell it to optimize toward. A single Performance Max campaign spanning a full catalog has one budget pool and one blended conversion value goal, and Google's own description of the campaign type is direct about the result: PMax "automatically funnels your budget toward the highest performing inventory and placements" inside that pool, not evenly across whatever you sell.
That is efficient from Google's side and opaque from yours. A high-margin line can get starved in favor of a high-volume, low-margin one, and the campaign-level ROAS can still look healthy the entire time, because the number is blended across both.
The same dynamic shows up over the life of an account, not just across a catalog. As a store's revenue mix shifts from mostly new customers toward mostly repeat buyers, a single ROAS-optimizing campaign has an easier time hitting its target by re-selling people who were already going to buy than by finding someone new. Nothing in the dashboard flags that shift happening. The fix is not better targeting inside the same campaign. It is a structural decision to isolate the budget that has to go toward prospecting from the budget that is allowed to chase warm, already-convinced traffic.
The layer that is actually safe to consolidate
Ad groups and keywords are a different story, and the instinct to resist consolidation here usually costs money rather than saving it.
A simple gut-check for whether a granular, exact-match structure is still earning its keep: exact match almost always carries a real CPC premium over broad match on the same underlying query, so that precision only pays for itself if the exact-match click converts enough better to cover the gap. Most accounts never run that comparison and keep granular structures out of habit, not evidence.
Google's own guidance on organizing Performance Max asset groups points toward a similar discipline at the campaign level: group assets by a clear theme, such as content category, language, or target audience, rather than splintering them into groups too narrow to have a real identity. A themed asset group with enough volume to actually learn from beats several thin ones competing for the same limited signal.
The practical rule: consolidate until you hit a reporting or budget boundary you actually need, then stop. The boundary, not the instinct to simplify, should decide where a campaign ends.
Naming conventions are structure, and they are where reporting quietly breaks
The part of account structure that never makes it into a "should I merge these campaigns" debate is naming, and it is the part most likely to corrupt a report six months from now.
A campaign named inconsistently with how your GA4 ecommerce tracking setup parses source and medium will still spend and convert correctly. It will also show up as its own, slightly different line in a channel report, splitting what should be one channel's performance across two or three rows nobody notices are the same thing until someone tries to reconcile total paid spend against total attributed revenue and the totals do not match.
The same ambiguity applies to conversion action ownership. If a Shopping campaign and a Performance Max campaign can both claim the same purchase without a clear rule for which one the business treats as the source of truth for that customer, your Performance Max reporting will always look slightly inflated relative to blended revenue, for reasons structure created rather than tracking.
Decide the naming convention and the conversion-action ownership rule before scaling spend, not after a founder asks why platform-reported revenue and the bank deposit disagree.
A short structure check before you touch anything
Before reorganizing an account, confirm the things most guides skip:
- Is brand demand isolated into its own campaign, so non-brand and PMax ROAS reflect prospecting only, not cheap brand capture?
- Does every campaign name follow one convention your GA4 channel grouping actually parses correctly?
- If a Performance Max campaign spans more than one margin tier, do you know how much of its reported value comes from your lowest-margin products?
- Is there a written rule for which campaign type owns credit for a conversion when more than one could plausibly claim it?
If any of those come back uncertain, a Google Ads audit is the faster path than guessing your way through a restructure, because it reconciles the account against GA4 and Shopify before you move a single campaign.
Deciding what your account actually needs
Skip the generic "best practice" template and work from what your store actually is.
A single-category store under meaningful daily spend needs the floor: one brand campaign, one non-brand campaign, one Shopping or PMax campaign. Anything more granular than that is probably broad-match-ad-group territory, not campaign territory, so consolidate it down.
A multi-category catalog with a real margin spread needs PMax split by category or margin tier, because a single blended campaign will hide which half of your catalog is actually funding the other half's acquisition cost.
An account past the point where returning customers outweigh new ones needs a deliberate, separately budgeted prospecting campaign that automation is not allowed to quietly defund in favor of warm retargeting, whatever the blended ROAS says.
An account about to scale spend materially needs the naming convention and conversion-action ownership rule settled first, because a structure that silently misreports at low spend will misreport at a much more expensive scale once budget moves.
None of this is about having more campaigns or fewer. It is about only adding a boundary where the business genuinely needs one, and never where "simpler looks tidier" is the only argument for merging two things that were reporting different answers. If your ads management partner cannot tell you which specific boundary in your account exists for a measurement reason versus a historical one, that is usually the first sign a structure review is overdue. A fixed-scope paid media audit is the fastest way to get that answer before a restructure, rather than after one goes wrong.
