How to Audit a Performance Max Campaign (When You Can’t See Inside It)

Performance Max is the campaign type most advertisers can least explain. Ask someone running Search campaigns how their budget is spent and they can point to keywords, match types, and search terms. Ask the same person about their Performance Max campaign and, more often than not, you get a shrug and “it’s doing fine, I think.” Google built PMax to run largely on automation — one campaign, one budget, and Google’s models decide how much goes to Search, Display, YouTube, Discover, Gmail, and Maps inventory. That convenience is exactly why so few people actually audit it.

The problem isn’t that Performance Max is bad. It’s that “set it and let it run” is a completely different discipline from “set it and check on it,” and most advertisers quietly default to the former. This post is a practical walkthrough of how to actually audit a live Performance Max campaign — where to look, what the reports are actually telling you, and what a reasonable audit cadence looks like — using the reporting tools Google Ads gives you today.

Why Performance Max Audits Look Different From Search Audits

A traditional Search campaign audit is a keyword-and-query exercise: pull the search terms report, look for waste, add negatives, adjust bids, repeat. Performance Max removes several of the levers that audit is built around. There are no ad groups, in the traditional sense — there are asset groups. There’s no keyword-level bidding. Negative keywords can only be applied at the account level or through a support request for brand exclusions, not added freely inside the campaign the way you would in Search. And because one campaign can serve across six different inventory types, “cost per conversion” at the campaign level tells you almost nothing about where that cost is actually coming from.

That’s the core issue: PMax collapses six different channels into one performance number. A campaign audit that only looks at the campaign-level dashboard is really just reading a summary — it can’t tell you whether your budget is going to high-intent Search queries or to YouTube views and Display placements that happened to get credited with a conversion. Auditing Performance Max well means deliberately pulling that summary apart.

Start With the Channel Performance Report

The first place to look is the channel performance report, found under Insights and reports on the campaign. This report breaks out impressions, clicks, cost, conversions, and conversion value by the individual channel — Search, Display, YouTube, Discover, Gmail, and Maps — instead of lumping them into one PMax number. Google has also added account-level channel reporting, so you can pull this across every Performance Max campaign at once rather than clicking into each one.

What you’re looking for here is concentration and drift. If a campaign you assumed was primarily a Search-intent play is quietly spending 40% of its budget on Display or YouTube inventory, that’s worth knowing — it changes how you should read the conversion numbers, and it may mean your creative assets (built for Search-style intent) are underperforming on channels they were never designed for. Channel mix isn’t static either; it can shift meaningfully month to month as Google’s models find new inventory, which is exactly why a one-time channel check isn’t enough — more on that later.

Pull the Search Terms Report — It’s Not Hidden Anymore

For a long time, one of the most legitimate complaints about Performance Max was that you couldn’t see the actual search queries triggering your ads. That’s no longer true. The search terms report for Performance Max is now available to all advertisers under Insights and reports, and it shows the real queries that triggered your search and shopping placements, along with the search categories Google groups them into and basic performance metrics.

Treat this the same way you’d treat a Search campaign’s search terms report: read it for relevance, not just volume. Are the queries triggering your ads actually connected to what you sell? Are there categories of query showing up that look like they belong to a different business entirely? Because you can’t add negative keywords directly inside the campaign, your main lever here is account-level negative keyword lists and brand exclusions — more limited than Search, but not nonexistent. If you see a consistent pattern of irrelevant queries, that’s a signal to either tighten your asset group’s audience signals and creative, or apply account-level negatives where the platform allows it.

Break Performance Down by Asset Group

Asset groups are the closest thing Performance Max has to ad groups, and they matter more than most advertisers give them credit for. Each asset group can target a different set of products, audiences, or themes, and asset group reporting lets you see performance at the individual asset group level, the campaign level, or across several campaigns at once. Asset reports can also be segmented by device, time, conversions, and network, and filtered by asset type and approval status.

If your campaign has multiple asset groups — say, one per product category or one per service line — a campaign-level average will hide the fact that one group is carrying the entire campaign while another is quietly burning budget with almost nothing to show for it. This is one of the most common things a PMax audit turns up: an asset group nobody has looked at since it was created, still running on outdated creative or a stale offer, still getting served, still spending.

While you’re in there, check asset-level performance too. Google labels individual assets — headlines, descriptions, images, videos — with performance ratings (low, good, best). An asset group full of “low” rated assets is a strong sign the creative needs to be refreshed, not that the audience or bidding strategy is wrong.

Check Where Your Ads Actually Showed

Google has extended the “Where ads showed” placement report to surface more detailed placement data for Performance Max, including the specific domains, apps, and YouTube channels your budget reached. This is the closest equivalent PMax has to a placement exclusion review on the Display Network, and it’s worth checking periodically for the same reason you’d check Display placements on any other campaign type: automated targeting occasionally finds inventory that technically drives a cheap conversion but does nothing for brand quality or long-term account health.

You’re not trying to micromanage every placement — that defeats the point of running Performance Max in the first place. You’re doing a periodic sanity check: does this list of domains and channels look like somewhere your brand should be showing up? If something looks clearly wrong (adult content, obvious clickbait networks, apps with no real audience overlap with your business), that’s a signal worth acting on.

Don’t Mistake Audience Signals for Targeting

A subtle source of confusion in Performance Max audits is what audience signals actually do. Unlike audience targeting in older campaign types, an audience signal in PMax is a starting hint for Google’s models, not a hard restriction on who can see your ads. Google can and will serve outside the signal you provide if its models believe a user is likely to convert. That’s by design, but it means an audit needs to check whether the signal is still doing useful work — pointing the algorithm in a sensible direction — rather than assuming it’s acting as a filter.

If an asset group’s audience signal was built from a customer list or in-market audience that’s now a year old, it’s worth refreshing with more current first-party data before assuming the campaign’s targeting problem is a bidding or creative issue. A stale signal doesn’t stop the campaign from spending; it just stops giving Google’s models a good starting point, and performance quietly degrades toward the account’s broader averages instead of your best customers’ profile.

Watch How Much of “Performance” Is Just Brand Search

One of the most consistent criticisms of Performance Max, raised repeatedly by PPC practitioners auditing live accounts, is that a meaningful share of PMax “performance” can simply be brand search that would have converted anyway — traffic from people already searching your business name, which the campaign happily takes credit for. This isn’t a flaw unique to PMax; it’s a broader attribution question in Google Ads. But because PMax pools Search inventory in with everything else, it’s easier for brand-driven conversions to get lost inside an otherwise-impressive top-line number.

The search terms report is your tool here too. When you review it, specifically separate out queries containing your brand name or close variants, and look at how the campaign performs without them. If a large share of your “efficient” conversions are coming from people who already knew your business, that changes how you should think about the campaign’s real incremental value — and it’s a much healthier way to evaluate whether PMax is actually finding you new customers or just efficiently closing out people who were coming anyway.

Build a Cadence, Not a One-Time Audit

Everything above is a snapshot. The real risk with Performance Max isn’t a bad initial setup — it’s drift that nobody notices because nobody is looking between audits. Channel mix shifts as Google’s models chase new signals. An asset group’s creative goes stale after a promotion ends but keeps serving anyway. A placement that was fine in month one starts showing up more heavily in month three. None of these show up as a dramatic, obvious failure in the top-line dashboard — they show up as a campaign that’s slowly getting less efficient while still technically “working.”

This is really the same problem across every part of a Google Ads account, and it’s the specific gap Growera’s continuous daily account monitoring is built to close — instead of PMax channel splits, asset group performance, and placement lists only getting reviewed when someone remembers to open the account, they get checked every day, so drift gets flagged while it’s still small rather than three months into a quietly declining campaign.

A Practical Performance Max Audit Checklist

Pulling the above together, here’s a working checklist you can run through on any live Performance Max campaign:

• Channel performance: Pull the channel performance report and check the split across Search, Display, YouTube, Discover, Gmail, and Maps. Does the mix match what you’d expect for this campaign’s goal?
• Search terms: Review the search terms report for relevance. Separate brand queries from non-brand queries and evaluate performance both ways.
• Asset group performance: Compare performance across asset groups rather than trusting the campaign-level average. Identify any group that’s underperforming or hasn’t been touched in months.
• Asset ratings: Check individual asset performance labels (low/good/best) and refresh anything consistently rated low.
• Placements: Review the “Where ads showed” report for Display and YouTube placements that don’t fit your brand.
• Audience signals: Confirm the audience signals attached to each asset group still reflect your actual customer profile, not whatever was set up at launch.
• Conversion settings: Double-check which conversion actions are feeding the campaign’s optimization — a bad or over-counted conversion action will quietly mislead every other metric in this list.
• Final URL expansion: If final URL expansion is turned on, check which landing pages Google has been directing traffic to. It can be a useful way to find high-converting pages you hadn’t prioritized, but it can also send traffic somewhere off-strategy if left unchecked.

None of these take long individually. The value comes from doing them regularly, not from doing any single one perfectly.

The Bottom Line

Performance Max isn’t a black box you have to accept on faith, and it isn’t a campaign type you can fully hand off to automation and forget about either. Google has steadily opened up more reporting — channel performance, search terms, asset group breakdowns, expanded placement visibility — specifically because advertisers pushed back on the “trust us” version of PMax. The reporting to run a real audit exists today; most accounts just aren’t using it.

The advertisers who get the most out of Performance Max treat it less like a fire-and-forget campaign and more like any other part of the account: something that needs a periodic, structured look at where the money is actually going, not just what the top-line conversion number says. Start with the channel and search terms reports, work down to asset groups and placements, and build that check into a routine instead of a one-time setup task. That’s the difference between a PMax campaign that’s genuinely finding new customers and one that’s just efficiently rediscovering the ones you already had.