High income consumer targeting is the practice of directing media spend at households above a stated household income threshold — typically $150K, $250K, or $500K+ — using data segments sold through DSPs, data marketplaces, and platform-native audience libraries. It is the most commonly used affluence filter in digital advertising, and in our experience running media for private aviation, luxury real estate, wealth management, private clubs, and premium DTC brands, it is also the least reliable one.
That is a strong claim, so here is the mechanism behind it. Almost nobody reports their income to a data provider. Income segments are therefore modeled — inferred from where you live, what you buy, what you browse, and what census tabulations say about households that look statistically similar to you. The label on the segment says "$250,000+ Household Income." The underlying data says "lives in a block group whose median income is high, and exhibits purchase behaviors correlated with that block group." Those are not the same statement, and the gap between them is where a great deal of luxury media budget quietly disappears.
This post covers how income segments are actually built, what precision to expect from each construction method, the four signals that consistently outperform income for high-consideration brands, and how to stack them into an audience that qualifies rather than merely widens.
How Household Income Segments Are Actually Constructed
There are four dominant methods, and they are not equally trustworthy.
Census block-group modeling. The provider maps a household's address to a census block group (roughly 600–3,000 people) and assigns the block-group income distribution to that household. Fast, cheap, near-total coverage — which is why it underpins the majority of large HHI segments. It is also the source of the most common defect in affluent targeting: a renter in a starter apartment two streets from a country club inherits the club members' income profile.
Credit-adjacent inference. Aggregated, anonymized, privacy-compliant credit-behavior signals feed a model that outputs an income band. Precision improves markedly over pure geography, and coverage is still broad.
Survey panel projection. A panel of tens of thousands of self-reporting consumers is weighted and projected onto tens of millions of devices. This is how the largest, cheapest income segments achieve their scale, and why they are the least accurate — the model is extrapolating hard, and self-reported income skews upward.
Transactional and registration evidence. Actual observed high-ticket purchases, property records, aircraft and vessel registrations, professional licensure. This is not income data at all — it is capacity data — and it is the most predictive input available for luxury advertisers.
Precision by construction method
These are the working ranges we use when evaluating a data partner. Precision means: of the users delivered in the segment, what share genuinely meet the stated threshold when validated against a client's known customer file.
| Construction method | Typical precision | U.S. scale | Recency | Relative CPM premium |
|---|---|---|---|---|
| Census block-group modeling | 35–50% | Very high (100M+) | Annual refresh | $0.50–$1.50 |
| Survey panel projection | 30–45% | Very high | Quarterly | $0.50–$2.00 |
| Credit-adjacent inference | 55–75% | High (60–90M) | Monthly | $2.00–$4.00 |
| Property and public record | 80–95% on the record | Moderate | Event-driven | $3.00–$6.00 |
| Observed transactional | 70–90% | Low (1–10M) | 30–90 days | $4.00–$9.00 |
Two things follow. First, scale and accuracy are inversely related — the biggest income segments in any DSP audience library are almost always the weakest, because scale is achieved through modeling. Second, the CPM premium on better data is trivially small next to the waste it prevents. Paying $4 more per thousand to double your qualified-reach rate is one of the easiest arbitrages in premium media.
High Income Is Not High Net Worth — and the Difference Decides Your Plan
This distinction is not academic. A $400,000-a-year household in a high-cost metro with two private school tuitions, a large mortgage, and no meaningful liquid position behaves nothing like a $180,000-a-year household with $6M in appreciated equity and no debt. Income measures inflow. Luxury purchases come out of balance sheet and discretionary confidence.
The industry has a name for the first group — HENRYs, High Earners Not Rich Yet — and they are the dominant population inside any large income segment. They are excellent prospects for aspirational premium DTC at $300–$1,500 price points. They are poor prospects for a $600,000 fractional jet card, a $4M second home, or a $10M advisory relationship.
When a client's product carries a lifetime value above $5,000 and a sales cycle longer than 30 days, income alone will produce a top-of-funnel that engages well and converts terribly. That specific signature — high CTR, high video completion, weak qualified-lead rate — is the diagnostic fingerprint of an income-only audience, and it is the most common problem we find in inherited media plans.
The Four Signals That Outperform Income
For high-consideration and luxury brands, we build the eligibility layer out of these instead.
- Property equity and ownership tier. County assessor and deed records are public, deterministic, and refresh on transaction events. Home value above a market-indexed threshold, ownership duration, mortgage-to-value position, and multiple-property ownership are all obtainable at the household level. Second-home ownership in particular is one of the strongest single predictors of luxury travel, private aviation, and club membership behavior we have measured.
- Investable-asset proxies. Modeled liquid-asset tiers from specialized wealth data providers are still models, but they are trained on materially better inputs than income models. Precision typically lands in the 60–80% range for the $1M+ investable tier, versus 35–55% for a comparable income segment.
- Category transaction evidence. Someone who has actually chartered a jet, purchased a $20,000+ watch, booked a $2,000-per-night suite, or bought a vehicle above $120,000 has demonstrated both capacity and willingness. Scale is small — often 1–5% of the size of a comparable income segment — but this is the layer that produces conversions.
- Occupational and entity identity. Business ownership, C-suite and partner titles, licensed professions with predictable earnings floors, and entity registrations. B2B-grade identity data is more accurate than consumer income data by a wide margin.
The stacking framework
The order matters, and getting it backward is the second-most-common defect we see.
- Layer 1 — Capacity (eligibility floor). Property equity, investable-asset tier, or occupational identity. This defines who can transact. Applied with AND logic, never OR.
- Layer 2 — Liquidity and life stage (qualification). Recent liquidity events, business sale signals, relocation, empty-nest transitions, inheritance-adjacent indicators. This defines who can transact now.
- Layer 3 — In-category intent (prioritization). Site visitation, comparison research, competitor engagement, first-party CRM activity. This defines who is actively considering.
A representative build for a private aviation client narrows from roughly 4.5 million capacity-qualified U.S. households, to about 300,000 with a liquidity or life-stage marker, to somewhere between 10,000 and 25,000 showing in-category intent in a given month. Each tier deserves different creative, different frequency, and different budget treatment — the intent tier can support $45–$70 CPMs on premium CTV inventory that would be indefensible against the top of the funnel.
Where Income Data Still Earns Its Place
None of this makes income segments useless. They are legitimately useful in three situations.
- As a suppression tool. Excluding the bottom income deciles is far more reliable than including the top ones, because the modeling error runs in the direction you can tolerate.
- As a scale extender for premium DTC. At $200–$1,500 price points, HENRY populations are the target, not the contamination. Income segments are appropriately matched to that economics.
- As a secondary qualifier layered onto a stronger primary. Property equity AND $250K+ HHI is a meaningfully tighter audience than either alone. Income works well in the second position and poorly in the first.
Activating High Income Audiences by Channel
Targeting precision is only half the equation — the delivery environment determines whether a qualified impression is worth anything.
Premium CTV. Household-level addressability on Disney+, Netflix, Prime Video, and premium FAST inventory makes CTV the strongest activation surface for capacity-qualified audiences, because the wealth signals above are household-level signals. Expect $38–$65 CPMs for tightly qualified affluent audiences, with completion rates of 92–97% on non-skippable 15s and 30s.
Programmatic display and video via private marketplace. Open exchange is where affluent targeting goes to die — not because the audience data fails, but because the inventory quality collapses. Running qualified audiences through curated PMPs and premium publisher direct deals typically lifts viewability from the high-50s into the 70–80% range.
DOOH. Geospatial targeting against private terminals, marinas, club corridors, and high-value residential zones is one of the few places where physical location is a genuinely strong wealth proxy rather than a weak one, because presence at an FBO or a private club is behavioral, not inferred.
Streaming audio and podcasts. Subscription-tier listeners skew meaningfully more affluent than ad-tier populations, and podcast audiences in business, finance, and golf verticals routinely index 150–220 against $250K+ households — often at lower effective cost than equivalent CTV reach.
Three Failure Modes and How to Diagnose Them
Geographic contamination. Symptom: strong delivery in high-income ZIPs, weak lead quality, and lead addresses clustering just outside the target neighborhoods. Cause: block-group modeling. Fix: replace geography-derived income with parcel-level property data.
The HENRY trap. Symptom: excellent engagement metrics paired with a qualified-lead rate below 1%. Cause: an income-only eligibility layer on a product priced against balance sheet. Fix: add a capacity layer with AND logic before anything else.
OR-logic stacking. Symptom: audience size grows when a qualifier is added. Cause: segments combined with OR instead of AND in the DSP audience builder. Fix: audit the Boolean logic in every line item; a properly stacked audience always shrinks.
The Validation Test Before You Spend
Before any income or wealth segment enters a plan, run three checks.
- Match-rate test. Hash the client's top revenue decile and ask the provider what share appears in the segment. Look for 25%+ overlap; treat anything under 10% as disqualifying.
- Composition audit. Ask what percentage of the segment is deterministic versus modeled. Thirty percent deterministic is our working floor for any primary qualifier.
- Population sanity check. Compare the segment size to the real-world population it claims to represent. Roughly 8.5 million U.S. households have $1M+ in investable assets; a "millionaire household" segment offering 40 million users is selling a model, not a population.
Work With Stillwater Media
Stillwater Media builds capacity-first affluent audiences for luxury and high-consideration brands, then activates them across premium CTV, curated programmatic, DOOH, and streaming audio with incrementality testing attached from day one. We take a limited number of engagements each quarter so that every audience is engineered rather than assembled. Apply to work with us →
