Performance Max Adds Household Income Exclusions: What It Means for Advertisers
Performance Max has always asked advertisers to trade control for automation — you hand Google your goals, assets and signals, and its models decide who sees the ad. On July 24, 2026, a small crack appeared in that black box. A new campaign setting, spotted in a European Performance Max account by paid-search specialist Thomas Eccel, lets advertisers exclude users by Google's estimated household income — a lever that existed elsewhere in Google Ads for years but never inside PMax.
It is not a confirmed global launch. It surfaced as a limited test, Google has published no announcement, and it may widen, change or vanish. But if it sticks, it hands PMax advertisers their first real audience-level exclusion since the campaign type launched. Here is exactly what was found, why it matters, and how to use it without undercutting PMax's automation.
What was actually spotted
The setting appears at the campaign level in Performance Max, under audience controls, and offers seven household-income buckets that can be switched off. Google's household-income data is modeled — an estimate stitched from account and behavioral signals, available only in select countries — not a figure users report. The available exclusion tiers are:
| Income tier | What excluding it does |
|---|---|
| Top 10% | Removes the highest-earning decile from targeting. |
| 11–20% | Removes the next band below the top decile. |
| 21–30% | Upper-middle income band. |
| 31–40% | Middle income band. |
| 41–50% | Lower-middle income band. |
| Lower 50% | Removes the entire bottom half of estimated earners. |
| Unknown | Removes users Google can't assign to any income bracket. |
Two details matter. First, it is an exclusion tool, not a targeting one — you subtract brackets rather than bid only into them, which keeps PMax's automated optimization running across everyone who remains. Second, the "Unknown" bucket is usually the largest and most volatile segment; excluding it can shrink reach far more than the label suggests, because a big share of impressions never gets a confident income estimate.
Why this is a bigger deal than it looks
Performance Max deliberately stripped out the granular controls advertisers had in Search and Display. You could feed it audience signals — hints about who to prioritize — but you couldn't hard-exclude a demographic. Household income exclusions break that pattern: they are a true campaign-level subtraction, not a suggestion the algorithm is free to ignore.
That is the same theme running through PMax's slow re-opening over the past two years — brand exclusions, account-level negatives, channel-level reporting, and now demographic exclusions. Google spent the launch phase selling PMax as a set-and-forget box; it has spent the maturity phase quietly giving advertisers back the knobs they complained about losing. Income exclusions are the most pointed of those knobs, because income maps so directly onto whether a click can ever convert. For teams already fighting rising Google Ads costs, cutting spend on audiences that structurally can't afford the product is one of the few efficiency levers PMax has ever offered.
Who benefits — and who should leave it alone
Income exclusions are a scalpel, not a default. They pay off when income is a genuine purchase signal and hurt when it isn't.
- Premium and luxury brands. High-ticket goods, luxury travel, private wealth and premium home services can exclude the lower income tiers to stop paying for clicks that rarely convert.
- Financial services. Wealth management, private banking and investment products often have hard suitability thresholds — excluding lower brackets aligns delivery with who can actually qualify.
- Automotive and big-ticket retail. New-car and premium-appliance advertisers can trim the bottom of the funnel where affordability collapses conversion rates.
- Value and budget brands — in reverse. Discount retailers and pay-in-installments products can exclude the top tiers to concentrate budget on price-sensitive shoppers who respond to their offer.
Who should be cautious: mass-market brands, anything with a wide price ladder, and lead-gen where the buyer's income doesn't predict intent. Income is a modeled estimate available only in some countries — exclude aggressively and you can starve PMax of the volume its bidding needs to learn, hurting performance more than the wasted spend you cut. This is a filter for advertisers with a clear affordability signal, not a universal setting to flip on.
How to use it without breaking PMax
If the control reaches your account, treat it like any other structural change to an automated campaign — deliberately, and with a way to measure the effect.
1. Confirm you have a real income signal
Before excluding anything, check whether income actually correlates with conversions in your data. If your top-of-funnel and bottom-of-funnel earners convert at similar rates, exclusions will just cut reach for no efficiency gain.
2. Start with one tier, not five
Exclude the single least-valuable bracket first — often "Lower 50%" for premium brands — and hold everything else. Sweeping exclusions across multiple tiers at once makes it impossible to tell which cut helped and which starved the algorithm.
3. Handle "Unknown" with extreme care
Because unattributed users are frequently the largest segment, excluding "Unknown" is the most aggressive move available and the easiest way to collapse volume. Test it in isolation, if at all, and watch impression share closely.
4. Give the algorithm time and measure against a baseline
Any exclusion resets part of PMax's learning. Note conversion rate, CPA and total conversions before the change, then judge the result over a full learning window rather than the first noisy days. If efficiency doesn't improve, reverse it — a narrower audience is only a win if it lifts return, not just tidiness.
The bigger picture for PMax advertisers
Whether this specific setting ships worldwide or not, the direction is clear: Performance Max is becoming less of a sealed box and more of a steerable one. That is good news for skilled advertisers and a trap for lazy ones — every new control is also a new way to over-constrain an algorithm that thrives on scale. The winners will be teams that add exclusions surgically and prove each one with data, the same discipline that separates efficient shopping campaigns from wasteful ones. If keeping up with PMax's constant stream of new levers is more than an in-house team can absorb, it strengthens the case for handing execution to a specialist — our take on when to use a white-label PPC agency weighs that trade-off. Either way, PMax in 2026 rewards advertisers who understand exactly which knobs to turn, and household income is the newest one on the board.
Frequently asked questions
Is household income exclusion officially live in Performance Max?
Not confirmed. It was spotted as a limited test in a European Performance Max campaign on July 24, 2026, and Google has issued no official announcement. It may expand, change or be withdrawn.
What income tiers can you exclude?
Seven segments: Top 10%, 11–20%, 21–30%, 31–40%, 41–50%, Lower 50%, and Unknown household income. You switch off the brackets you don't want; the campaign keeps optimizing across the rest.
Will excluding income tiers improve my Performance Max results?
Only if income genuinely predicts conversions for your product. Because Google's income data is modeled and available in limited countries, aggressive exclusions can shrink reach and starve PMax's bidding of the volume it needs to learn — test one tier at a time against a baseline.
Should I exclude the "Unknown" income segment?
Be very careful. Unattributed users are often the largest segment, so excluding "Unknown" is the most aggressive option and can drastically cut impressions. Test it in isolation and monitor impression share before committing.
