Google PMax: Brand Exclusions and Channel-Level Reporting Are Live

What Google has actually added to PMax

Over the past three years Google has progressively reversed the original PMax design philosophy of "let the AI handle it," adding the granular controls, exclusions, and reporting transparency advertisers had been asking for since the campaign type's 2022 launch.

Dive deeper into your channel performance with a granular view of metrics for all your channels including Search, YouTube, Discover, Gmail, Display Network, Search partners and Maps in the channel distribution table. Google Ads - blog.google

The four buckets of change matter most. First: controls. Brand exclusions (live since 2024) let you block specific brand terms - your own or competitors' - from triggering PMax on Search and Shopping inventory. Account-level negative keywords cap at 1,000; campaign-level caps at 10,000. Demographic exclusions and device targeting are now available. Second: reporting. Channel-level performance, full search terms, asset-level metrics with impressions/clicks/cost, and a downloadable channel distribution table. Third: targeting. Search themes cap raised from 25 to 50 per asset group. URL-contains rules for feed-based campaigns. Fourth: diagnostic tooling. Channel performance diagnostics flag specific issues (low Maps coverage from missing store locations, low Search investment from landing-page relevance, etc.).

The pace has accelerated. The 2024 Google Ads recap listed channel performance reporting and brand suitability transparency as headline launches. The January 2025 PMax features post added campaign-level negative keywords to all advertisers, brand exclusions for different formats in retailer campaigns, demographic exclusions, and device targeting. The May 2025 channel performance reporting expansion added search-terms reporting parity with Search and Standard Shopping. In aggregate this is a different campaign type than the one most accounts were last fully audited against.

Timeline of the transparency updates

These features rolled out as a sequence of separate launches across 2023, 2024, and 2025. The relevant dates and what each launch unlocked:

• Announced: First wave: 2023 (account-level negative keywords) (Google Ads Help Center + Google Ads blog)
• Rollout begins: Brand exclusions GA: 2024 (per Google's New brand settings post) for global, all PMax advertisers
• Full rollout: Channel-level reporting + search terms reporting + campaign-level negatives: 2025 (per Jan 2025 PMax features blog)
• Enforcement begins: ongoing - the 2025 highlights add Waze inventory, retention goals, demographic/device targeting

Who benefits most

The accounts that gain the most from the new controls are the ones where PMax was consuming the budget of other campaign types (especially branded Search) and the ones where channel-mix opacity was driving conservative spend allocation. The new reporting closes both gaps.

Two caveats. Hotel Ads campaigns inside PMax for travel goals do not support negative keywords or brand exclusions per Google's documentation. And the channel performance report has known reporting quirks - the troubleshooting guide explicitly notes that video conversion windows default to three days and that PMax "is designed to maximize your overall campaign performance across all channels, not just the performance of any single one," which means channel-by-channel optimization is directional, not absolute.

What to do this week

Priority order: pull the data you now have access to (search terms, channel distribution, asset-level metrics), then add the controls (negative keywords, brand exclusions, demographic and device targeting), then iterate on the channel mix. Skipping the data step and going straight to controls is the most common audit failure.

1. Pull the channel distribution table for the last 90 days. Open each Performance Max campaign and download the channel distribution table. Look at the split across Search, YouTube, Discover, Gmail, Display, Search Partners, and Maps. If your CPL or CPA goal is lead-gen intent and YouTube/Display is consuming 40%+ of spend, that is the audit signal. The diagnostics section will flag specific causes (landing page relevance, missing store locations, etc.).
2. Pull a full PMax search terms report. Search terms reporting in PMax now matches Search and Standard Shopping campaign granularity per Google's 2025 blog post. Export the search terms, sort by spend, and identify (a) high-spend low-converting queries to negative-out, (b) high-volume branded queries that should be moving through branded Search instead of PMax, and (c) intent-shift queries that suggest you need an asset-group split.
3. Apply brand exclusions for your own brand and direct competitors. For most ecommerce setups: exclude your own brand on Search text ads but keep Shopping (the retail pattern Google explicitly documents). For lead-gen: exclude your brand entirely if you have a separate branded Search campaign. Add 3-10 direct-competitor brand names if you have evidence PMax is bidding on their queries (search terms report from action 2).
4. Build out the negative keyword stack. Use the 10,000 campaign-level cap and 1,000 account-level cap. Per Google's 2025 recap, campaign-level negative keyword lists are now applicable to PMax, which means a shared exclusion list (e.g., a luxury retailer's "cheap, bargain, discount" list) can apply to multiple campaigns at once. Most accounts we audit are running fewer than 200 negative keywords on PMax campaigns that should have 1,000+. Tactical guide in our negative keyword strategy for 2026.
5. Audit asset-level performance and search themes. Asset reporting now includes impressions, clicks, and cost alongside conversions per Google's 2025 update. Look at which asset types are driving spend without performance. Search themes cap is 50 per asset group (up from 25); the search themes "usefulness indicator" added in 2025 tells you which themes are actively informing the AI's targeting and which are dead weight.

What to do this quarter

The strategic shift: PMax has moved from "set it and forget it" to "set it, monitor it, and steer it." Per Glenn Gabe and Marie Haynes' kitchen-sink framing on Google updates, PMax accounts that win in 2026 are not the ones running the most automation - they are the ones using the new controls to remove waste channels, block cannibalization, and keep the AI focused on actual incremental conversion paths.

Restructure the PMax + Search relationship

PMax is designed by Google to be additive to a separate Search campaign per the company's own positioning ("Search and Performance Max are the Ads Power Pairing"). If you are running PMax in isolation as a single account-wide campaign, brand exclusions plus a dedicated branded Search campaign will almost always restructure the spend more efficiently than letting PMax cover branded and non-branded queries with one campaign. The exception: very small accounts where the extra structure adds more overhead than it saves.

Set a recurring channel-distribution audit cadence

Build a monthly review where you pull the channel distribution table, the search terms report, and asset-level metrics for every PMax campaign over $5K monthly spend. Track channel mix month-over-month. If YouTube or Display share is creeping up and your goal is lead-gen, that is the early-warning signal to tighten demographic targeting or device exclusions before performance slips. Most accounts we audit do not have any structured PMax review cadence - they react to performance drops instead of preventing them.

What we're seeing in real accounts

Note: the patterns below are aggregated from audits we've run for ecommerce and lead-gen accounts in 2025-2026. The dominant finding: most accounts are still using the 2022 PMax mental model (opaque black box, minimal controls), which means the new transparency features are sitting unused. The accounts that adopt them within the first audit cycle typically see efficiency gains in the 10-25% CPA range.

Counterexample: a lead-gen account in the B2B services vertical that had already implemented brand exclusions and was running a structured PMax + Search campaign pair saw smaller gains (~6% CPL improvement) from the new transparency features. The lesson is that PMax efficiency is bounded by the structure around it - the transparency updates help accounts that were structurally inefficient, not accounts that were already running PMax with discipline.

What we're still watching

Four open questions are driving how we sequence PMax audit work for the next two quarters.

• AI Mode Network reporting: Whether Google adds AI Mode Network as a separate reportable channel inside PMax. Currently PMax spend on AI Mode surfaces appears to roll up under Search, which makes channel-mix analysis incomplete. Our breakdown of the differences is in Google Search Network vs AI Mode Network.
• Search themes usefulness signals: Whether Google expands the search themes "usefulness indicator" (added 2025) into a numeric score. Currently it is a directional signal; a quantified score would let us prune dead themes faster.
• Feed-only PMax controls: Whether feed-only PMax (no asset group) gains additional granular controls beyond URL-contains rules. The feed-only mode is the closest analog to legacy Shopping campaigns and gets disproportionate use in ecommerce accounts.
• AI Max integration with PMax: Whether the AI Max for Search campaigns feature evolves to be controllable from within PMax (effectively a single campaign type covering both). Our writeup on AI Max is in AI Max for Search campaigns.