Google Ads Account Structure: Best Practices for 2026

If you have hired a Google Ads consultant or an agency in the last few years, there is a decent chance the first thing they did was rebuild your account structure before touching a single bid. That is not busywork. Account structure is the scaffolding everything else sits on: budgets, Quality Score, negative keywords, reporting, and increasingly, how well Google's automated bidding systems understand what you are trying to sell. Get the structure wrong and every optimization on top of it fights against itself.

This guide covers how to think about Google Ads account structure in 2026, when Performance Max and broad match have changed a lot of the old rules, and what a sane Google Ads campaign structure actually looks like for a business spending anywhere from a few thousand to six figures a month. If you would rather have someone build and manage the structure for you, that is exactly what our Google Ads management service does. This article is for the "how does this actually work" crowd.

Why Account Structure Still Matters

Every couple of years someone declares account structure dead because Google's machine learning "figures it out anyway." It is true that Google's systems lean more heavily on signals like landing pages, audience data, and conversion history than they did five years ago, and it is true that over-segmenting an account into hundreds of ad groups with three keywords each is worse than it used to be, because it starves each ad group of the conversion volume the algorithm needs to learn.

But "less granular" is not the same as "no structure." Structure still controls three things automation cannot fix for you: how your budget is allocated across products or services, how your messaging matches search intent, and how cleanly you can read your own reporting later. A flat account with fifteen keywords stuffed into two ad groups will spend efficiently in Google's eyes and still fail your business, because you will not be able to tell which service line is actually generating revenue versus which one is just eating budget.

The Three Levels: Account, Campaign, Ad Group

A Google Ads account has three structural layers, and each one exists to control a different variable:

• Account: billing, conversion tracking, shared negative keyword lists, and audience data live here. Most businesses need exactly one account. Multiple accounts under one business almost always create more reporting headaches than they solve.
• Campaign: this is your budget and bid strategy boundary. Each campaign gets its own daily budget, its own bid strategy (Maximize Conversions, Target CPA, Target ROAS, manual CPC), and its own network settings (Search, Display, Shopping, Performance Max, YouTube). If you want to control spend independently between two things, they need to be separate campaigns.
• Ad group: this is your message-to-search-term match. Ad groups hold a tight cluster of keywords (or, for Performance Max, a theme) and the ads that speak directly to that cluster. This is where relevance and Quality Score live.

How to Structure Campaigns Around Intent and Budget

The most durable way to structure campaigns is around the business decisions you actually need to make, not around keyword themes for their own sake. Three splits earn their complexity almost every time:

By funnel stage. Separate high-intent, ready-to-buy search terms (your brand name, "near me," "pricing," "buy") from research-stage terms ("what is," "how does," "best"). These convert at wildly different rates and deserve different budgets, different bid strategies, and honestly different landing pages. Blending them into one campaign means your bid strategy is optimizing toward an average that describes neither group well.

By product line or service, when margins or sales cycles differ. A roofing company running both emergency repair and full re-roof campaigns should split them: repair leads close in days, re-roof leads take weeks and are worth far more per conversion. One shared budget and one Target CPA across both will systematically starve whichever service has the higher CPA, even if it is the more profitable one.

By geography, only when budgets or bid strategies genuinely need to differ by location. Location targeting can usually be handled inside a single campaign using location bid adjustments. Split into separate geo campaigns when you are testing different budgets per market, running seasonal promotions in one region and not another, or reporting separately to different franchise owners or business units.

For B2B accounts specifically, this is also where lead quality starts to show up in your structure. If one campaign is consistently producing more Sales Qualified Leads per dollar than another, that is a signal to shift budget toward it, not just a reporting footnote.

Ad Group Structure: How Granular Should You Go?

This is where most accounts go wrong in one of two opposite directions: either everything is dumped into two or three catch-all ad groups, or the account is split into hundreds of micro ad groups left over from the single-keyword-ad-group era.

A workable middle ground for most Search campaigns is 5 to 15 tightly related keywords per ad group, all sharing the same core intent, with 2 to 3 responsive search ads written specifically for that theme. "Tightly related" means a searcher looking at any keyword in the group would be satisfied by the same ad and the same landing page. If you find yourself writing an ad that has to hedge its language to cover two different keyword themes, that is the signal to split the ad group.

Is SKAG Still Worth It in 2026?

Single Keyword Ad Groups, one keyword per ad group, each with its own exact-match version, were the dominant structure philosophy roughly a decade ago. The logic held up when match types were rigid and Quality Score rewarded keyword-to-ad text matching almost word for word.

It mostly does not hold up anymore, for a specific and testable reason: Google's close-variant matching (even on exact match) and the shift toward Smart Bidding mean each single-keyword ad group now gets a fraction of the impression volume it used to, and Smart Bidding needs conversion volume per ad group to learn well. Splitting an account into hundreds of one-keyword groups spreads that volume so thin that most groups never leave the "Learning" phase, and Smart Bidding performs worse, not better, than it would with slightly broader groups.

There are still narrow cases where near-SKAG structure earns its keep: very high-value, low-volume B2B terms where you want granular control over ad copy and bids on a handful of head terms specifically. For the majority of accounts, the 5-to-15-keyword themed ad group described above outperforms it in practice.

Structuring Around Performance Max and Broad Match

Performance Max campaigns do not have ad groups in the traditional sense, they have asset groups, and Google explicitly discourages segmenting them too finely for the same volume-starvation reason described above. The Performance Max playbook we cover elsewhere goes deeper on this, but the structural implication for the rest of your account is worth calling out here: when Performance Max sits alongside standard Search campaigns, keep a clear line between them using negative keywords and audience signals so they are not bidding against each other on the same searches.

Broad match keywords, combined with Smart Bidding, work best in their own dedicated campaigns or ad groups rather than mixed in with phrase and exact match terms. Mixing match types in one ad group makes it very hard to tell which match type is actually driving performance when you review the data later, because Google's reporting attributes the click to whichever match type "won" the auction, and that can shift week to week for the same underlying search term.

Common Structural Mistakes That Waste Budget

• One campaign, one budget, for a business with multiple service lines of different profitability. This is the single most common issue we see in accounts that "aren't converting": the budget is quietly being consumed by the lower-margin service because the bid strategy cannot tell the difference.
• Keeping Search and Shopping in the same campaign type mindset. Shopping and Performance Max campaigns pull from your product feed, not your keyword list, so the discipline that matters there is feed quality and product grouping, not ad group granularity. See the Shopping setup guide if that is your situation.
• Letting legacy ad groups pile up. Accounts that have been running for years often carry dozens of paused or near-zero-volume ad groups from strategies that were abandoned but never cleaned out. They rarely hurt performance directly, but they make the account unreadable for anyone trying to audit it, including you.
• No shared negative keyword lists. Without them, the same irrelevant search terms get excluded campaign by campaign, manually, forever, instead of once at the account level.

When and How to Restructure an Existing Account

Restructuring is disruptive: Smart Bidding strategies and Quality Score history reset to some degree whenever you move keywords into new ad groups or campaigns, so it is not something to do on a whim. A few situations genuinely warrant it:

• You are adding a new product or service line that does not fit cleanly into the existing campaign logic.
• Budget allocation has become impossible to control because unrelated offers are sharing one campaign budget.
• The account was built years ago around SKAG or another outdated philosophy and conversion volume per ad group is too thin for Smart Bidding to perform well.
• You are inheriting an account from a previous agency or freelancer and cannot make sense of the existing logic well enough to report on it confidently.

When you do restructure, do it incrementally where possible: move one campaign at a time, keep the old structure paused rather than deleted for a few weeks as a fallback, and expect a short dip in stability while Smart Bidding relearns. According to Google's own guidance on campaign structure, giving each campaign or ad group time to exit the learning period after a change, generally one to two weeks with consistent conversion volume, matters more than moving fast.

A Simple Framework You Can Apply This Week

If you are auditing an existing account or building a new one, this order of operations covers most businesses without over-engineering it:

1. List your actual business decisions first: which service lines need independent budgets, which need independent bid strategies, which need separate reporting for a client or stakeholder. That list becomes your campaign split, nothing more.
2. Inside each campaign, group keywords by shared search intent, not by loose topical similarity. Aim for ad groups a human could write one honest, specific ad for.
3. Build one shared negative keyword list at the account level for terms that are never relevant to any campaign (competitor names you don't want, "jobs," "free," "DIY," and similar), and campaign-specific lists for terms that are only irrelevant in that context.
4. Separate match types and campaign types (Search, Shopping, Performance Max, Display) rather than blending them, so your reporting later tells you the truth about what is working.
5. Leave it alone long enough to get real data before judging it. A structure needs conversion volume to prove itself, not a week.

Frequently Asked Questions