Account Structure & Tools

Granular Account Structure vs Hagakure Structure

In short: These sit at opposite ends of how much an account trusts automation versus manual control. Granular structure splits campaigns and ad groups along business lines and often keeps tighter match types, so a human retains the levers. Hagakure consolidates into a small number of broad ad groups and hands relevance decisions to Smart Bidding and broad match. Both can be correct depending on the business - they just optimize for different things. If the business genuinely needs separate budgets or bid strategies by line, granularity wins; if the goal is maximum algorithmic performance from limited human oversight, Hagakure wins.

By the AdFlint research team · Fact-checked against current Google and Meta platform behavior · Last reviewed July 2026

Granular Account Structure

Many narrow campaigns and ad groups split by keyword, product, geography, or match type to maximize manual control over budget and messaging.

Splitting the account buys you precise budget caps, separate bid strategies, and reporting you can read at a glance. It genuinely helps large accounts with distinct margins, regions, or inventory that must not share spend. The failure mode is applying it to a small account: each container gets a fraction of the conversions, automated bidding never gathers enough data, and you spend your week managing structure instead of performance.

Full definition

Hagakure Structure

Google's consolidation philosophy: few campaigns, broad keywords, and responsive ads, so machine learning sees pooled data instead of scattered fragments.

Originating from Google Japan, Hagakure argues that automated bidding performs better when traffic concentrates in a small number of high-volume ad groups rather than spreading across many thin ones. Advertisers adopt it when moving to Smart Bidding and broad match. The misunderstanding is reading it as an instruction to delete all structure. You still need separation where budgets, geographies, or margins genuinely differ, and negatives still do real work.

Full definition

Side by side.

The differences that actually change what happens in your account.

	Granular Account Structure	Hagakure Structure
Number of containers	Many - split by product, region, margin, or match type.	Few - a small number of broad ad groups per campaign.
Match type stance	Often tighter, phrase or exact, to keep control over which query triggers which ad group.	Broad match by default, trusting the algorithm to find relevant queries.
Who holds the levers	The advertiser - budget caps, bid strategy, and targeting differ container by container.	The algorithm - Smart Bidding decides how to allocate spend within each broad container.
Conversion data per container	Lower per container, since volume is split across many splits.	Higher per container, since traffic pools into fewer ad groups.
Negative keyword role	Moderate - mostly separates one split from another.	Load-bearing - the main thing keeping broad match on-topic.
Best fit	Larger accounts with real budget or targeting distinctions across lines.	Accounts with enough budget per ad group and a mature negative list, willing to trust the algorithm.
Failure mode	Splitting a small account until no container has enough data to bid intelligently.	Broad match drifting into irrelevant queries when negatives lapse, or ad groups too broad for a coherent landing page.

What actually separates them.

Granular structure adds containers to increase control; Hagakure removes containers to increase data density - they move account structure in opposite directions.

Granularity typically pairs with tighter match types to preserve the manual control it's built for; Hagakure typically pairs with broad match to let the algorithm do the work granularity would otherwise do by hand.

Under granularity, a human decides which budget or bid strategy applies to which slice of the business; under Hagakure, Smart Bidding decides how to spend within a broad pool the human has stopped subdividing.

Negative keywords do light cleanup work in a granular account but carry the main weight of relevance control in a Hagakure account.

A large account with genuinely distinct product margins usually needs some granularity that pure Hagakure would erase; a small account drowning in thin containers usually needs the consolidation Hagakure represents.

Which one should you use?

Use Granular Account Structure when

Different products or regions carry different margins and can't share a budget.
You need a different bid strategy for part of the account, like brand versus prospecting.
A stakeholder needs performance reported and budgeted by a specific business line.
You don't yet trust broad match enough to hand over targeting decisions the way Hagakure requires.

Use Hagakure Structure when

You have a mature, actively maintained negative keyword list and are comfortable letting broad match find new queries.
Your budget per ad group is large enough to gather meaningful data even after consolidating.
You're rebuilding an account from scratch and want to lean into Smart Bidding rather than replicate manual control.
Your business doesn't have distinct margins or regions that genuinely need separate budgets.

Common questions.

Can a large account be both granular and Hagakure-structured?

Only in parts, and it takes deliberate design - you might keep granular, tightly-matched campaigns for a high-margin product line while running broad-match, Hagakure-style campaigns for the rest of the catalog. Applying both philosophies to the same campaign at the same time doesn't really work, since they pull in opposite directions.

Which one performs better?

Neither wins universally - it depends on whether the business has real distinctions worth protecting with separate budgets or bid strategies. If it does, granularity's control usually pays for itself; if it doesn't, Hagakure's data pooling usually outperforms unnecessary splits.

I split my account by region and performance got worse. Is that a sign I should go Hagakure instead?

Possibly, if the regional split didn't actually need separate budgets or bidding and just diluted the data each campaign had to learn from. Try merging the regions that don't have a real reason to be separate before abandoning granularity entirely - you may only need to undo part of the split.

Or stop choosing between them.

AdFlint picks the setting, writes the ads, and keeps optimizing inside the Google and Meta accounts you already own.

Try AdFlint free See how it works

Related comparisons

All comparisons