The Millionaire Household Targeting Playbook

Here's the uncomfortable truth most agencies won't say out loud: there is no "net worth" field in The Trade Desk, DV360, or Amazon DSP. Millionaire household targeting digital campaigns don't work by flipping a switch that says "show this only to millionaires." Net worth is not observable, not self-reported at scale, and not something any ad platform collects directly. Every vendor that claims otherwise is selling you a modeled proxy — the only question is whether that proxy is built well.

For luxury brands, this distinction isn't academic. The U.S. has roughly 8 million households with a net worth over $1 million excluding primary residence, according to data cited by wealth research firms — a little over 6% of all households. Spread across the country's 330 million-person population and the noise of a national programmatic buy, that 6% is a needle in a haystack unless your targeting stack is built specifically to isolate it. This playbook breaks down how Stillwater Media actually finds and reaches these households: the data layers that work, the ones that are mostly marketing claims, and how to combine them into a media plan that doesn't waste two-thirds of its impressions on people who will never buy.

Why "Millionaire" Isn't a Targeting Category

When a CMO asks for millionaire household targeting, what they usually mean is "reach people who can afford and are inclined to buy what we sell." That's a behavioral and economic profile, not a single data point — and conflating the two is the single most common mistake in affluent media planning.

Consider two households both worth roughly $2 million. One is a 68-year-old couple who paid off their home decades ago, holds most of their wealth in retirement accounts, and spends conservatively. The other is a 42-year-old dual-income professional household with a large mortgage, two leased vehicles, and active discretionary spending on travel, dining, and home renovation. Both are "millionaire households" by net worth. Only one of them is a relevant audience for most luxury brands — and it's usually the second one, despite having less total wealth on paper.

This is why Stillwater doesn't lead affluent campaigns with net worth as the primary filter. We lead with propensity to spend in-category, and use wealth indicators as a qualifying layer underneath that.

The Data Stack: What Actually Works

Reaching millionaire households digitally requires stacking several data sources, because no single source is both accurate and scalable. Here's how the layers break down, roughly in order of reliability:

• 1. First-party transaction and CRM data. If your brand already has any customer data — even a few thousand records — this is your highest-signal source. Match rates back to identity graphs run 60-80% for well-hygiened files, and the resulting lookalike models inherit real purchase behavior rather than modeled assumptions.
• 2. Modeled household income and net worth (third-party data providers). Companies like Experian, Acxiom, and Epsilon maintain household-level models built from public records (property values, mortgage data), survey panels, and credit-adjacent signals. These models are directionally useful — a household modeled in the top 1% of income is meaningfully more likely to be affluent than one in the 50th percentile — but they're probabilistic, not verified. Expect 70-85% directional accuracy at the household level, lower at the individual level.
• 3. Geographic and property-value signals. Home value, lot size, and neighborhood median income correlate strongly with household wealth, and unlike income models, property data is largely a matter of public record. Geo-fencing affluent zip codes, specific neighborhoods, or even individual high-value parcels (via geofencing technology) is one of the most underused tactics in luxury media — and one of the most precise.
• 4. Behavioral and contextual signals. What someone does online is often a better predictor than what a model says about their bank account. Searching for private jet charter, browsing $200K+ vehicle configurators, or engaging with premium real estate listings are stronger purchase-intent signals than a modeled income decile — these are the "luxury buyer intent signals" that should anchor any affluent campaign.
• 5. Premium content and platform context. Where someone spends attention is a wealth proxy in itself. Audiences on Disney+, premium financial news sites, golf and aviation content, and first-class travel apps over-index for affluence in ways that are both measurable and contextually brand-safe.

Building the Segmentation Model

A well-built millionaire household targeting strategy layers these sources rather than relying on any one. Here's the practical sequence we use:

• Start with first-party data, if any exists. Even a customer list of a few hundred high-LTV buyers can seed a lookalike model that outperforms any off-the-shelf affluent segment.
• Layer in modeled wealth and income data as a filter, not a target. Use it to exclude clearly mismatched audiences (e.g., suppress household income under $100K for a $150K+ vehicle campaign) rather than to define the entire audience.
• Add geographic precision. Identify the zip codes, counties, or specific neighborhoods where your historical customers concentrate, and weight media delivery toward those geographies — even within a national buy.
• Layer behavioral intent on top. This is the difference between "probably affluent" and "actively in-market." A household that is both modeled-affluent and currently researching your category is exponentially more valuable than either signal alone.
• Validate with holdout testing. Run a geo-holdout or audience-holdout test to confirm that the layered segment actually converts at a higher rate than a broader buy — don't assume the model works just because the inputs sound sophisticated.

Channel-by-Channel: Where Millionaire Households Actually Spend Time

Different channels offer different precision-to-scale tradeoffs for affluent targeting. Here's how we think about the mix:

No single row in this table is "the answer." The campaigns that perform best for our clients — JetLinx, FLY Exclusive, Financial Independence Group — combine high-precision, low-scale channels (DOOH, native) to build the seed audiences and creative context, with high-scale programmatic and CTV to extend reach against modeled lookalikes of those seed groups.

Platform Considerations: Where the Data Actually Connects

The data layers above only matter if they connect to the platforms doing the buying. In practice, the connective tissue is identity resolution — matching a household-level data record (income model, property data, behavioral signal) to a device or browser that a DSP can actually bid against.

The Trade Desk offers the deepest third-party data marketplace access of the major DSPs, with direct integrations to providers like Experian, Epsilon, and Eyeota — making it the most flexible environment for layering modeled wealth data with behavioral segments. DV360 leans more heavily on Google's own first-party signals (affinity and in-market audiences), which are strong for behavioral intent but offer less granular household-level wealth data than The Trade Desk's marketplace. Amazon DSP has a unique advantage for affluent targeting in retail-adjacent categories: Amazon's purchase history data is arguably the strongest real-world spending signal available at scale, since it reflects actual transactions rather than modeled estimates — though it's limited to Amazon's own data and doesn't extend to non-shopping behavioral signals.

For CTV specifically, Disney+, Netflix, and Prime Video each offer their own first-party household data (subscription tier, content consumption, in some cases purchase history via Amazon), which can be layered with the DSP-side targeting described above when buying through their respective ad platforms or private marketplace deals.

The practical implication: a millionaire household targeting strategy that works well on one DSP may need to be rebuilt — not just copied — for another, because the underlying data partnerships and identity graphs differ meaningfully. This is one of the most common reasons campaigns that perform well on one platform underperform when "the same audience" is pushed to a second platform without adjustment.

Affluent Lookalike Modeling: Getting It Right

Affluent lookalike modeling is one of the most misused tools in luxury media. The mistake we see most often: agencies build a lookalike model off website visitors or social engagers, without filtering for which of those visitors actually converted or had high order value. The result is a lookalike model of "people who clicked," not "people who buy" — and it scales the wrong audience efficiently.

The fix is straightforward but requires discipline: seed lookalike models exclusively from your highest-value converters — ideally customers in the top quartile of lifetime value — and refresh the seed list quarterly as new conversions come in. A 50,000-person seed list of actual high-LTV buyers will produce a tighter, more efficient lookalike than a 500,000-person list of general website visitors, even though the smaller list feels less "statistically robust" on paper.

Common Mistakes in Millionaire Household Targeting

• Mistake 1: Treating modeled net worth as ground truth. Modeled wealth data is directionally useful at the aggregate level but noisy at the individual level. Use it to weight delivery, not to gate it entirely — overly aggressive net-worth filters often suppress reach into the 15-20% range where the model simply lacks data, even though many of those households are genuinely affluent.
• Mistake 2: Ignoring household composition. A $5 million net worth household with three kids in college has very different discretionary spending patterns than a $5 million household of empty-nesters. Layering household composition data (available from most data providers) sharpens targeting considerably for categories like luxury travel, real estate, and premium automotive.
• Mistake 3: Optimizing to CTR instead of downstream value. Affluent audiences often have lower click-through rates than general audiences — they're more selective with their attention — but convert at significantly higher values when they do engage. Campaigns optimized purely to CTR systematically de-prioritize affluent segments in favor of cheaper, higher-CTR — and lower-value — audiences.
• Mistake 4: Skipping frequency management. Millionaire households are a small enough universe that frequency caps matter enormously. Without disciplined capping, a campaign can burn 40%+ of its budget on repeat impressions to the same few thousand households within days.
• Mistake 5: No holdout group. Without a holdout, there's no way to know whether the "affluent" segment is actually outperforming a broader buy — or whether the premium CPMs are buying nothing but a narrative.

How to Measure Success

Standard digital KPIs — CTR, CPC, even ROAS in the short window — often understate the value of millionaire household targeting because the sales cycle for luxury purchases typically runs 30-90+ days. The metrics that matter more:

• Conversion value per converter, not conversion rate alone — a smaller number of high-value conversions can outperform a larger number of low-value ones
• Geo-holdout incrementality — comparing matched markets with and without the affluent-targeted campaign to isolate true lift
• Audience overlap with known high-LTV customers — using match-rate analysis to confirm the targeted audience genuinely resembles your best existing customers
• Frequency distribution — ensuring impression delivery isn't concentrated on a small repeat-exposed group

The Bottom Line

Millionaire household targeting digital campaigns succeed when they stop looking for a magic targeting checkbox and start treating affluence as a layered signal — first-party data where it exists, modeled wealth and geography as a filter, and behavioral intent as the activation trigger. The brands that get this right aren't necessarily spending more; they're spending more precisely, and they're validating that precision with holdout testing rather than assuming the targeting works because the data vendor's sales deck said so.

If your current media plan treats "affluent" as a checkbox in a DSP rather than a layered strategy, it's worth a conversation. Stillwater Media accepts a limited number of new client engagements per quarter.