On July 24, 2026, paid search specialist Thomas Eccel posted a screenshot from a European PMax account showing something that hadn't been there before: a campaign-level setting to exclude users by household income bracket. Top 10%, 11–20%, 21–30%, all the way down to Lower 50%, plus an "Unknown" bucket. No press release. No Google documentation. Just a live control sitting in campaign settings, waiting to be toggled.
That's either an opportunity or a trap, depending on how you treat it.
Why This Matters for Pipeline Quality
PMax has always been a black box with limited demographic controls at the campaign level. Income-based exclusions existed in Search, Display, and YouTube campaigns across roughly 20 countries where Google models purchasing power. But PMax? You got audience signals (directional, not strict) and whatever Google's automation decided to do with them. That was it.
If this feature rolls out broadly, it adds a lever that demand gen teams have wanted: the ability to trim segments that consistently produce low-quality leads. Think premium SaaS with $50k+ ACV, financial services, high-end professional services. In those categories, income can correlate with buying authority and budget access. Excluding brackets that never convert past MQL could reduce wasted spend and sharpen pipeline quality.
But the word "could" is doing a lot of work in that sentence.
The Precision Problem
Household income is the least precise of Google's core demographic categories. It's inferred, not observed. Google doesn't see bank statements; it models purchasing power from aggregate signals. That means false positives and false negatives are baked in. Exclude the Lower 50% bracket and you might cut a segment that includes a VP at a Series C startup whose household income data is stale, miscategorized, or simply absent.
The "Unknown" bucket deserves special attention. In B2B SaaS, where audience segments are already narrow, excluding Unknown income could shrink reach more than expected. Google's demographic coverage isn't uniform across geographies or user profiles. Cutting Unknown is tempting (it feels like cleaning up noise), but the volume hit can be disproportionate.
The trade-off you're accepting: tighter targeting at the cost of reach in a channel where reach is already constrained by automation decisions you can't fully see.
How to Test Without Breaking Anything
If the setting appears in your account, resist the urge to exclude three brackets at once. Here's a tighter approach.
The hypothesis (make it falsifiable): If we exclude [one income bracket] from PMax Campaign X, then qualified lead rate (measured at booked-call or SQL stage) will increase by ≥15% over 3 weeks, because we're removing a segment that historically converts at the top of funnel but drops off before pipeline.
Setup: Pick a single PMax campaign with enough volume to read results (minimum 50 conversions per week). Exclude one bracket. Run for 14–21 days.
What to measure: Don't rely on in-platform CPA alone. That number will almost certainly improve (you're removing volume). The real question is downstream: did booked calls, qualified pipeline, or revenue per lead actually move? Validate in your CRM. Match Google Ads click IDs to post-conversion outcomes.
Guardrails: Set a stop-loss. If total conversion volume drops more than 25% without a corresponding quality lift at the SQL stage, revert. Monitor daily for the first week.
What not to over-interpret: A CPA decrease paired with a volume drop isn't proof of quality improvement. It's arithmetic. You need the CRM data to confirm the signal.
The Bigger Operational Question
This feature is undocumented. Google hasn't announced it. It appeared in one European account and got shared on LinkedIn. That's the entire evidence base. Building a major strategy around it would be premature.
What isn't premature: having a process for detecting and evaluating new PMax controls as they surface. Google ships features into accounts without announcements regularly. If your team doesn't have a monthly account-audit cadence that checks for new settings, you're either missing levers or (worse) running with defaults you didn't choose.
Income exclusions aren't a substitute for strong audience signals. First-party data, Customer Match lists, high-value converter signals still do more heavy lifting in PMax than any single demographic exclusion. Treat income brackets as a secondary lever, not a primary strategy.
Thomas Eccel spotted a setting in one account on a Thursday in July. Whether it becomes a standard PMax control or disappears in the next update, the real test isn't whether the feature exists. It's whether your measurement infrastructure can tell you if it actually moved pipeline.
