Competitor Friction
Why is Performance Max so opaque?
You are not imagining it. Performance Max reports at a coarser level than the campaign types it replaced, so several questions you used to answer in the interface now have to be answered outside it.
Some reporting has improved: there is channel performance reporting and a search terms report. What is still hard to read is which combination of creative, audience and product carried the spend, and how that maps to your own contribution. The fix is to read what is published and reconcile it to your orders.
Channel
Performance Reported
Search
Terms Reviewed
Weekly
Video Reports
Orders
Reconciled To Reporting
The Problem
What PMax hides from you
These are reporting limits of the campaign type rather than faults in your account. Each one makes a commercial decision harder to take from the interface alone.
Asset Groups With No Visibility
You cannot see which creative, audience, or product combination drove a sale. PMax bundles everything together into a single attribution layer.
Partial Search Term Data
Performance Max does have a search terms report, but terms below Google's privacy thresholds are not shown and queries are not a bidding lever inside the campaign.
Branded And Non-Branded Demand Blended
Branded and non-branded demand report as one number by default, so the split is hard to read. A large branded share is a structural question to investigate, not proof that the conversions were non-incremental.
Audience Signals Ignored
You add custom audiences, remarketing lists, and in-market segments. Google treats them as optional suggestions and ignores them when it finds easier conversions elsewhere.
Worked Example
Hypothetical worked example - £45k/month PMax spend
What PMax reporting shows you versus what our visibility layer reveals. Same account, same data, completely different conclusions.
Blended View: One Number
Split View: Branded And Non-Branded Read Separately
The Insight
These figures are illustrative, not a client result. The 5.8x blended number is arithmetically consistent with a 10x branded segment and a 3x non-branded segment sitting inside it. At a 20% contribution margin the non-branded £81,000 of revenue produces £16,200 before ad spend, which is £10,800 short of the £27,000 spent, before overhead. Reading the split tells you where the money went; it does not by itself prove the branded conversions were non-incremental. Branded share is not incrementality, and that question needs a designed test.
Our Approach
How do you fix Performance Max visibility?
We do not accept Google's defaults. We build systems that extract the data you need to make informed commercial decisions.
- Read the reporting Google publishes - Channel performance, search terms, asset group and listing group data in one place. See Performance Max reporting: channels, search terms and profit.
- Separate branded and non-branded demand - Make the split readable; proving incrementality needs a designed test.
- Segment by commercial intent - Build asset groups around objectives, not product taxonomy.
- Establish weekly Loom reports - Get walkthrough explanations of what happened and why.
Reporting Layer You Can Read
We pull the reporting Google does publish - channel performance, search terms, asset group and listing group data - into one place, reconciled to your own orders. Scripts save time; they cannot surface data Google does not publish.
Brand Separation Architecture
We structure campaigns so branded and non-branded demand report separately, which makes the split readable. Reading it is not the same as proving incrementality: that needs a designed test.
Asset Group Segmentation
We build asset groups around commercial intent rather than product taxonomy, and where margin bands need different targets we separate them into their own campaigns, because targets and budgets are campaign-level settings.
Weekly Transparency Reports
Every week you receive a Loom walkthrough explaining what happened, what changed, and why it matters to your bottom line - not just a dashboard screenshot.
Sector-Specific
PMax opacity by sector
Different sectors face different PMax visibility challenges. The fix is never generic - it must account for your industry's specific dynamics.
Fashion & Apparel
The Challenge
PMax combines product images with lifestyle assets, so it is hard to read which visual approach drove a sale. Where returns are material, platform-reported revenue overstates what the business kept, and the return rate is something to measure in your own data rather than assume.
The Fix
Group products into asset groups by commercial intent, and set targets and budgets at campaign level, because tROAS targets and budgets are campaign settings rather than asset-group settings. Where returns are material, adjust reported conversion values so optimisation reflects kept orders rather than placed orders.
Beauty & Skincare
The Challenge
Subscription products and gateway SKUs create LTV dynamics PMax cannot see unless you feed them in. It optimises toward the conversion values you send it, so a first-order-only value makes a gateway product look weaker than it is.
The Fix
Feed LTV-adjusted conversion values into PMax. Separate gateway products into their own campaign where the target needs to differ, since the target applies to the campaign, and keep asset groups organised around intent.
Home & Living
The Challenge
Wide price ranges (£15 cushion vs £2,000 sofa) mean PMax gravitates toward high-conversion, low-margin products. Your expensive items get starved of budget.
The Fix
Create separate PMax campaigns by price band with different tROAS targets calibrated to each band's break-even point. The cushion at 55% margin needs a very different target than the sofa at 18% margin.
How do you fix Performance Max opacity and see what's actually working?
TLDR: Read the reporting Google does publish, reconcile it to orders, and test rather than infer causation.
We read the reporting Google publishes - channel performance, search terms, asset group and listing group data - reconcile it against your own orders and contribution, and separate branded from non-branded demand so the split is readable. Asset groups are built around commercial intent rather than product taxonomy, and where margin bands need different targets they are separated into their own campaigns, because targets and budgets are campaign-level settings. Where a causal question matters, it is answered by a designed test, not by dashboard inference.
- Branded share of PMax:
- Account specific: measure it, do not assume it(JudeLuxe)
How much of Performance Max spend goes to branded search?
TLDR: The branded share of PMax is account specific. Isolate brand traffic and measure it rather than assuming an industry average.
It varies enormously by brand, and we do not publish an average because we cannot evidence one. What we can say is that the branded share is usually larger than the account owner expects, and it is invisible until you isolate it. Without brand exclusions, PMax can function as an expensive branded search campaign, taking credit for traffic you would have received organically or through cheaper branded campaigns. Brand isolation architecture is what lets you measure it in your own account rather than guess.
Frequently Asked Questions
PMax visibility and transparency
Want to see what PMax is hiding?
Book a PMax visibility review. We will read the reporting Google publishes for your account, show how branded and non-branded demand split, and reconcile it against your own orders.
Book PMax Visibility AuditPrefer to see how the work is run day to day? Read how our Performance Max agency team structures asset groups, brand separation, and campaign-level targets. The day to day method sits on our Performance Max management page.