Why Performance Max Makes Bad Accounts Look Average

Performance Max has a reputation for improving results. In many cases, it does. But there is a quieter pattern that gets less attention: PMAX can also mask problems.

A badly structured, poorly optimised account will often show improvement when PMAX is added. Not because PMAX fixed the problems, but because PMAX hides them.

The Smoothing Effect

PMAX is designed to find conversions wherever they exist across Google's inventory. It reallocates budget dynamically. It optimises toward the conversion goal it is given.

When layered on top of a struggling account, this smoothing effect can produce modest improvement. Conversions increase. ROAS stabilises. The panic recedes.

But the underlying issues remain. Product-level margin problems persist. Budget allocation is still suboptimal. The account is still structurally flawed. PMAX has simply papered over the cracks.

Why This Matters

The danger is not that PMAX fails. The danger is that PMAX succeeds just enough to prevent the harder work from happening.

If PMAX brings a 2.5x ROAS account to 3x, it looks like progress. But if the account should be achieving 4x with proper structure and segmentation, PMAX has actually locked in underperformance.

The modest improvement becomes the new baseline. The structural problems are forgotten. The ceiling is lowered, and no one notices because the floor was raised.

The Attribution Complication

PMAX also complicates attribution. It runs across channels, including Search, Shopping, Display, YouTube, and Discovery. It claims conversions that may have happened anyway. It cannibalises branded search. It takes credit for remarketing.

This is not fraud. It is how the system works. But it means PMAX performance cannot be read at face value.

A Performance Max audit separates incremental value from claimed value. It asks: what would have converted without PMAX, and what is PMAX actually adding?

The Segmentation Problem

PMAX works best when given clean signals. When the product feed is well-structured, when asset groups are thoughtfully segmented, when the algorithm knows what success looks like.

But most PMAX implementations are not like this. They are broad, catch-all campaigns with minimal segmentation. The algorithm optimises toward volume because volume is easier. Margin is ignored because margin is not visible.

This is where bad accounts start to look average. PMAX finds the path of least resistance. That path is rarely the most profitable one.

The Agency Incentive

PMAX is appealing to agencies because it requires less granular management. One campaign, broad targeting, algorithmic optimisation. The reporting is simpler. The time investment is lower.

This is not cynicism. It is resource allocation. Agencies have margins too.

But it means PMAX is often implemented as a shortcut rather than a strategy. The account improves from bad to average, and that improvement is reported as success.

The harder question, whether the account could be genuinely good with proper structure, is rarely asked.

What Good PMAX Looks Like

Good PMAX implementation does not replace structure. It requires structure.

It requires a feed that is segmented by margin, not just category. It requires asset groups that reflect commercial priorities. It requires exclusions that prevent cannibalisation. It requires reporting that separates PMAX performance from overall account performance.

This is what we focus on with Performance Max management. Not using PMAX as a fix, but building the foundation that allows PMAX to perform.

The Question to Ask

If your PMAX campaigns are delivering moderate results, ask: is this the ceiling, or have we just stopped looking?

If PMAX was added without addressing underlying account structure, the answer is probably the latter.

Average is not the goal. Average is what happens when the harder work does not get done.