The algorithm needs fewer campaigns. You still need more answers.

Consolidate the signal, not the thinking

Granular account structures feel disciplined because every product, device, audience, and theory gets its own boundary. The problem begins when those boundaries divide a limited supply of conversions into pieces too small to guide automated bidding. At that point, the structure is not creating control. It is creating a collection of under-informed campaigns.

The useful question is not how many campaigns an account should have. It is which decisions genuinely need a separate budget or bid target, and which divisions merely starve the system of signal.

That distinction matters because a neat account map can hide a weak learning environment. Ten carefully named campaigns that each produce only a few conversions may give the operator more switches to touch, but not enough information for any one switch to work well. Consolidation is valuable when it concentrates the evidence the bidding system is allowed to use.

The boundary is important: the material supports conversion density as the decision principle, not one universal threshold for every account. Budget, sales cycle, conversion quality, and the decisions that need protection still shape the final structure.

PMax can spend the signal faster than you can explain it

This is where a preference for less complexity becomes dangerous shorthand. A consolidated PMax campaign can spend faster than the Search campaigns it replaced because it has more places to spend. That may be useful reach. It may also be inventory that would never have appeared in the old Search setup. The blended return does not tell you which one happened.

So do not diagnose faster spend by staring harder at the campaign total. Inspect channel performance and asset-group placement reports. Find where the money went before deciding whether consolidation improved the account, widened it, or simply made waste harder to see. If a placement consumes spend without producing the outcome the campaign exists to buy, exclude it surgically before cutting the whole budget.

This is not an argument against PMax. It is an argument against treating the same target as proof that two campaign types are doing the same job. Consolidation gives the system more data and more freedom. The reporting work must become sharper in return.

Fewer campaigns does not mean a flatter business

Campaign count and useful internal structure are different decisions. A business may need fewer bidding boundaries while still preserving product ranges, offer differences, or other meaningful distinctions inside ad groups and assets. Collapsing campaigns should not collapse the information architecture marketers need to read performance and make merchandising decisions.

A sound consolidation pass asks four things:

1. Which outcomes require a genuinely separate budget or bid target?
2. Which campaigns lack enough conversion activity to support themselves?
3. Which product or offer distinctions can remain visible inside a consolidated campaign?
4. Which PMax channels and placements are spending without earning the result being optimized?

Those questions prevent two opposite mistakes: splitting the account until every campaign is hungry, or merging it until the business becomes invisible inside one efficient-looking number.

Give up the control that was imaginary

The old promise of granular structure was that more levers meant more control. When every lever works from weak data, that control is mostly theatre. But the answer is not to hand the account to automation and stop asking what it bought.

Consolidate where fragmentation weakens the signal. Preserve structure where a real business decision depends on it. Then use placement and channel evidence to audit the freedom you gave the algorithm. The goal is not the smallest possible account. It is the smallest account that still tells you enough to act.