How to Use Google's Channel Performance Report to Finally Understand PMax

Performance Max has always done one thing very well: spending your budget. What it has never done well is telling you where. Google's Channel Performance Report is the most meaningful transparency update PMax has received since launch — and most advertisers are either ignoring it or misreading it.

This guide covers exactly what the report shows, where it misleads you, and how to use it to make decisions that actually improve campaign performance. No guesswork, no vague advice — just what the data means and what to do with it.

What the Channel Performance Report Actually Shows

Before this report existed, advertisers running Performance Max campaigns had no way to see how budget was split across Google's network. You saw total impressions, total spend, total conversions — all pooled together into a single campaign row. Shopping, Display, YouTube, Gmail, and Discover were treated as one undifferentiated pool.

The Channel Performance Report breaks that pool apart. It shows you, for any given date range, how your PMax campaign actually allocated its budget and generated conversions across individual channels. It also introduces two critical segmentation dimensions that have never been available in native reporting before:

• Ad type segmentation — separating feed-based ads (generated from your product data) from asset-based ads (generated from the headlines, images, and videos you upload)
• Conversion event type — separating click-through conversions (genuine conversion actions following an ad click) from view-through conversions (conversions attributed to an ad impression that was never clicked)

These two dimensions together answer the questions that have plagued PMax advertisers since the format launched: Is my campaign actually driving revenue, or is it reporting impressions as conversions? Is my product feed doing the work, or is Google using my creative assets in ways I didn't intend?

The report is available at both the account level and the individual campaign level. The account-level view gives you a summary table across all PMax campaigns simultaneously. The campaign-level view adds the Sankey flow diagram and more granular channel-by-channel data tables with export options.

How to Access and Navigate the Report

You find the Channel Performance Report under Campaigns → Insights & Reports → Channel Performance in the Google Ads interface. It is currently labeled as beta in most accounts. If you do not see it, check that your account has at least one active Performance Max campaign with sufficient historical data — accounts with very low traffic may not have it surfaced yet.

At the account level, the report displays a table with one row per PMax campaign. Each campaign row is expandable to show channel-level breakdowns. The columns show standard performance metrics: impressions, clicks, conversions, conversion value, and cost. By default, ROAS and CPA are not included — you need to add these manually through the Columns button under the Conversions section.

Adding ROAS and CPA per channel is not optional if you want the report to be actionable. Without these columns, you can see that YouTube generated 40,000 impressions and 12 conversions, but you cannot tell whether those conversions cost $3 or $300 each. The report is dramatically more useful once you have cost-per-conversion and conversion value visible at the channel level.

The campaign-level view adds the Sankey diagram above the data table. Clicking into a specific campaign from the account overview takes you here. The Sankey shows a flow from channels to ad types to conversion goals, which can be useful for understanding the general structure of how your campaign operates — but as covered in a later section, the proportions in the diagram are not accurate and should not be used for budget decisions.

The report supports full CSV and Google Sheets export. For any serious analysis — calculating CTR per channel, building a ROAS heatmap, comparing click-through vs view-through by ad type — you will want to export the raw table and work with it outside the interface.

Feed Ads vs Asset Ads: The Most Important Segmentation

The "Ads Using Product Data" segmentation — which separates feed-based ads from asset-based ads — is the single most practically useful feature in the Channel Performance Report. It answers a question that has been unanswerable in standard Google Ads reporting for years.

Feed ads are generated dynamically from your product feed. They include Shopping ads in Search, dynamic remarketing display banners, and catalog-based video formats on YouTube. Their performance is entirely dependent on the quality of your product data: title accuracy and keyword inclusion, image quality, price competitiveness, availability status, and completeness of custom labels and attributes.

Asset ads are generated from the creative assets you upload into your asset groups: headlines, descriptions, images, logos, and video files. These power responsive search ads, responsive display ads, and standard YouTube video ad formats.

Why does the split matter? Because the diagnosis for underperformance is completely different depending on which ad type is responsible:

• If feed ads are underperforming, the fix lives in your product feed — title optimization, adding GTINs, improving image quality, correcting pricing errors, completing missing attributes
• If asset ads are underperforming, the fix lives in your creative — weak headlines, generic descriptions, low-quality images, missing video assets, or assets that don't match the audience intent at each channel

Before this segmentation existed, optimizing PMax required guesswork about which layer of the campaign was responsible for results. Now you can see directly whether your feed is outperforming your creative or the reverse — and allocate optimization effort accordingly. For e-commerce accounts running both Shopping and creative-based display, this segmentation typically reveals major imbalances that would otherwise go undetected for months.

View-Through Conversions: The Hidden Inflation Problem

The conversion event type segmentation — separating click-through conversions from view-through conversions — addresses what has been one of the most significant sources of PMax result inflation since the format launched.

A view-through conversion is recorded when a user sees an ad impression but does not click it, and then converts on your site within a defined attribution window (typically 1 day for Display and 30 days for video). Google counts these in your conversion total by default, and they appear in your ROAS calculations as if they were the result of active ad engagement.

The problem is that most view-through conversions would have happened anyway. A user who sees a display banner while reading an article, ignores it, and later navigates directly to your site to buy something they already intended to buy — that is a view-through conversion that your PMax campaign is claiming credit for. On average, 40% of PMax-reported conversions fall into this category.

The Channel Performance Report lets you separate these two types for the first time in native reporting. You can now answer questions like:

• What percentage of my reported conversions are view-through? If it is above 30–35%, your ROAS figure is likely overstated
• Which channels are generating the highest view-through conversion rates? Display and YouTube are the most common sources
• Are specific asset groups or ad types disproportionately reliant on passive attribution? This can indicate the campaign is reaching audiences who were already converting rather than driving new demand

Understanding this split is essential for accurate campaign measurement. If your reported ROAS is 6x but 45% of conversions are view-through, your true paid-driven ROAS may be closer to 3–4x. Decisions about budget allocation, campaign scaling, and bid strategy targets should be based on click-through conversion data, with view-through treated as a supplementary signal rather than a primary performance metric.

The Sankey Diagram: What to Use and What to Ignore

The Sankey flow diagram in the campaign-level view shows the path from channels (Search, Display, YouTube, etc.) through ad types (feed or asset) to conversion goals. It is visually striking and intuitively appealing — the width of each flow appears to represent the volume of activity through that path.

The proportions are not accurate. This is not a minor caveat. The diagram regularly shows channels that represent a small fraction of actual spend or impressions as visually dominant flows, while channels that handle the majority of budget appear comparatively thin. Making budget decisions based on the diagram's visual proportions would lead you to incorrect conclusions about where your campaign is actually spending.

The Sankey is useful for one purpose: understanding the structural topology of your campaign at a glance. If you want to know whether your PMax campaign routes traffic through Shopping feed ads into purchase conversions, or through YouTube asset ads into lead form conversions, the Sankey shows you the connection pattern clearly. It tells you which paths exist — not how significant each path is.

For any quantitative analysis — which channel has the best ROAS, which ad type is driving the most conversions, how the split between click-through and view-through varies by channel — use the data table exclusively. Export it, add your calculated columns, and build your analysis from numbers rather than from the visual representation.

Google is aware of the diagram's limitations and has acknowledged that improvements to the visualization are in development. Until those improvements arrive, treat the Sankey as an orientation tool and nothing more.

Decoding What Each Channel Label Actually Means

Google's channel labels in the report use platform-level names that don't directly map to the ad formats you may be familiar with from campaign setup. Understanding what each label actually refers to in practice is necessary for interpreting the data correctly.

Search — Feed ads in Search are Shopping ads, showing product images, prices, and titles drawn directly from your feed. Asset ads in Search are responsive search ads and dynamic search ads, generated from your uploaded headlines and descriptions.

Display — Feed ads in Display are dynamic remarketing ads and dynamic prospecting display formats, generated from product data and shown to users on Google Display Network placements. Asset ads in Display are responsive display ads, assembled from your uploaded images, logos, headlines, and descriptions.

YouTube — Feed ads on YouTube include catalog-based video formats that pull from product data — some of these formats are not prominently documented by Google, which can make YouTube feed performance surprising when you first see it segmented. Asset ads on YouTube are standard in-stream, bumper, and non-skippable video formats using the video files you uploaded.

Discover and Gmail — These channels use asset-based formats exclusively. Discover shows card-style ads in the Google Discover feed; Gmail shows expandable ads in the Promotions tab. Both use images and copy from your asset groups.

This decoder matters because the same channel can serve very different ad types, and the optimization levers for each are completely different. A Display channel with poor performance could mean bad dynamic remarketing (fix the feed) or bad responsive display ads (fix the creative) — the channel label alone doesn't tell you which.

How to Use This Report to Improve PMax Performance

Reading the report is only the first step. The value comes from connecting what you see in the channel data to specific changes in campaign structure, creative, or measurement.

Identify channel imbalances and their causes. If Display is consuming 35% of your budget but delivering 8% of your click-through conversions, that imbalance needs investigation. The first question is whether the Display spend is feed-based or asset-based — because the fix is different. Feed-based Display waste often means your dynamic remarketing is reaching too broad an audience; asset-based Display waste often means your creative is not performing well in display formats. You cannot reduce Display spend directly, but you can influence it by improving Shopping feed quality (which pulls more budget toward the higher-performing Search channel) and by improving asset quality scores.

Separate real performance from attributed performance. Export the table, filter for click-through conversions only, and calculate your true click-driven ROAS by channel. Compare this to your overall reported ROAS. If the gap is larger than 20–25%, your bidding targets are set against inflated numbers and your campaign is likely underinvesting in channels that drive genuine demand while over-crediting passive impression channels.

Use channel data to prioritize feed and creative work. If Shopping feed ads are your highest-performing format by click-through ROAS but their volume is lower than you expect, invest in feed optimization — title rewriting, adding GTINs, improving custom label segmentation for your bidding strategy. If asset ads on YouTube have high impression volume but low conversion contribution, audit your video assets and consider whether you have provided enough video variation for the algorithm to test effectively.

Connect channel data to CRM revenue for full picture accuracy. Platform-reported ROAS from the Channel Performance Report still carries all the normal caveats of Google Ads attribution. Combining channel-level data with your CRM-tracked revenue — using UTM parameters built by channel and ad type — gives you 2–3x more reliable ROAS signals than platform data alone. This is the approach we use in our growth strategy engagements to make scaling decisions that compound rather than waste.

For context on how PMax fits into a broader paid search strategy in the current environment, see our breakdown of what's actually working in PPC in 2026 and our guide to advanced Google Ads techniques for accounts that have already mastered the fundamentals.