Luxury Intent Signal Targeting: Which Signals Actually Predict a Purchase

Luxury intent signal targeting is the practice of identifying affluent consumers who are actively moving toward a high-value purchase — not merely those wealthy enough to afford one. The distinction matters more than any other variable in premium media planning, because the overwhelming majority of what gets sold as "luxury audience targeting" is demographic proxy data dressed up in behavioral language.

Wealth is a qualifier. Intent is a signal. Confusing the two is why a private aviation operator can spend $200,000 reaching a perfectly composed HHI $500K+ audience and generate almost no qualified inquiries. Everyone in that audience could buy. Almost none of them were going to.

This piece covers the signal taxonomy we use at Stillwater Media, the predictive strength of each signal class, how decay windows should govern activation timing, and how to build a composite score that outperforms any single vendor segment.

Why Demographic Targeting Fails for High-Consideration Purchases

Consider the arithmetic. There are roughly 8.5 million U.S. households with investable assets above $1 million. In any given quarter, perhaps 15,000 of them will seriously evaluate a fractional jet program. That is an in-market rate of about 0.18%.

If your targeting is purely wealth-based, 99.8% of your delivered impressions land against people who are not going to act — regardless of how good the creative is. You have not bought a bad audience. You have bought a correct audience at the wrong moment, which produces the same outcome and costs the same money.

Intent signal targeting attacks the denominator. The goal is not to find richer people; it is to find the subset already in motion, and to concentrate spend against them during the window when media can actually influence the outcome.

The Luxury Intent Signal Taxonomy

Not all intent signals are equal. We classify them into five tiers by predictive strength, and we weight them accordingly in scoring models. The table below reflects the lift ranges we typically observe when a signal class is isolated against a matched wealth-only control.

Signal Class	Example Signals	Predictive Lift vs. Wealth-Only	Typical Decay Window	Scale
Declared intent	Quote request, brochure download, tour booking, calculator use	8–20x	7–21 days	Very low
Deep research	Multi-session comparison browsing, spec-sheet views, pricing-page revisits	4–9x	14–45 days	Low
Category browsing	Editorial reading, review sites, category landing pages	2–4x	30–60 days	Moderate
Life-event triggers	Liquidity event, relocation, business sale, marriage, retirement	3–7x	60–180 days	Low–moderate
Adjacency behavior	Ownership or engagement with correlated luxury categories	1.4–2.5x	90–180 days	High

Two things stand out. First, predictive strength and scale move in opposite directions — the strongest signals are the rarest, which is why intent targeting must be layered rather than singular. Second, decay windows vary by nearly an order of magnitude, and this is the variable most brands ignore entirely.

Decay Windows: The Variable Almost Everyone Gets Wrong

A declared-intent signal is worth an enormous amount for about ten days and very little after thirty. A liquidity-event signal is worth relatively less on day one and considerably more on day sixty, because the buyer needs time before they begin shopping.

Most programmatic setups treat all audience segments with the same recency logic — typically a flat 30-day lookback — which simultaneously over-serves stale declared intent and under-serves life-event signals that have not yet matured. The correction is straightforward but rarely implemented:

• Declared intent: activate within 24–48 hours, heavy frequency for 7 days, taper hard by day 21.
• Deep research: activate within 72 hours, sustained moderate frequency for 30–45 days.
• Category browsing: steady low frequency across 60 days, used primarily to build the retargeting pool.
• Life-event triggers: delayed activation, ramping from day 14 to day 90 as the buyer enters an active phase.
• Adjacency behavior: always-on baseline at minimal frequency; this is prospecting fuel, not conversion media.

In practice, restructuring decay logic alone typically improves cost per qualified inquiry by 20–35% before any change to inventory, creative, or bid strategy.

Building a Composite Intent Score

No single signal source is sufficient. Vendor segments are noisy, first-party signals are sparse, and contextual signals lack identity. The approach that works is a composite score that blends sources and weights them by demonstrated predictive value against your own closed-won data.

The construction sequence we use:

1. Establish the ground truth set. Pull 12–24 months of closed customers from the CRM, including deal value and time-to-close. This is the only honest benchmark for whether a signal predicts anything.
2. Onboard and resolve identity. Match CRM records to an identity graph so you can observe what those customers did in the 90 days before conversion. Without identity resolution, everything downstream is guesswork.
3. Back-test candidate signals. For each available signal — first-party, vendor, contextual, and platform-native — measure its presence rate in the converter set versus a matched non-converter set. Anything below roughly 1.3x lift should be discarded, not discounted.
4. Assign weights and build tiers. Convert lift ratios into a weighted score. We typically bucket into four activation tiers rather than a continuous score, because programmatic activation works better with discrete audiences.
5. Set tier-specific bids and frequency. The top tier can justify 3–5x the CPM of the bottom tier. Bidding uniformly across an intent-scored audience wastes the entire exercise.
6. Re-fit quarterly. Signal predictive value decays as data partnerships change and consumer behavior shifts. A model that is not re-fit within six months is degrading.

The output is not a segment. It is a scoring system that assigns every reachable affluent household a probability of being in-market, and lets you spend accordingly.

First-Party Signals Are the Highest-Value Input — and the Most Underused

The strongest intent signals for a luxury brand are almost always ones the brand already owns and does not activate. In audits, we routinely find:

• Pricing and configurator page visits sitting in analytics but never piped to an activation audience.
• Sales-call disposition data in the CRM that distinguishes "not now" from "not ever" — enormously valuable for suppression and re-engagement, almost never used for media.
• Email engagement depth (clicks on specific content types, not just opens) that reliably separates browsers from evaluators.
• Event and showing attendance for real estate, automotive, and private club categories, which is close to declared intent and typically lives in a spreadsheet.
• Service and ownership records that identify existing customers whose upgrade cycle is approaching.

Activating these requires a clean data pipeline and a consented identity resolution path, which is real work. But first-party signals outperform purchased segments by a wide margin — in our client programs, first-party-derived audiences typically deliver 2.5–6x the conversion rate of comparable third-party intent segments, at no incremental data cost.

Activating Luxury Intent Signal Targeting Across Channels

A composite score is only useful if the channels you buy can act on it. Addressability varies substantially, and the activation plan has to respect those limits rather than assume uniform precision.

Premium CTV supports household-level onboarded audiences through most major supply partners, making it the best channel for high-tier intent activation. Match rates on a clean, hashed CRM file typically land between 55% and 75% against household graphs — high enough that top-tier intent audiences can carry meaningful frequency. This is where the strongest signals should be spent.

Programmatic display and native offer the finest-grained control: tier-specific bid multipliers, per-audience frequency caps, and creative variation by score band. Addressability is lower post-cookie, generally 40–60% in Chrome-dominant environments, so plan reach against the matched subset rather than the nominal segment size.

Streaming audio and podcasts support demographic and contextual targeting well, but household-level intent activation is inconsistent across publishers. We generally treat audio as a mid-tier channel — worth activating research-stage and life-event signals, less suited to declared intent that demands rapid response.

DOOH cannot address individuals, but it can be informed by intent data through geographic weighting. If your top-tier intent households cluster in eleven ZIP codes, the DOOH network should follow that geography rather than a generic affluent-district buy.

Paid social offers strong onboarding for custom audiences but poor transparency. Use it for suppression and mid-tier retargeting, and never let its self-reported conversions arbitrate whether your intent model is working.

Creative should change with the score

The most consistent uplift we see from tiered activation comes from varying creative by intent tier, not just bid. Top-tier households should receive specific, offer-adjacent creative with a clear next step. Bottom-tier prospecting audiences should receive brand and category education, because a hard conversion ask against a household 90 days from readiness produces avoidance rather than action. Running one creative across all tiers forfeits most of the advantage the scoring model created.

How to Evaluate a Premium Consumer Data Partnership

Third-party intent data still has a role, particularly for prospecting beyond your own footprint. But the market is full of segments whose methodology does not survive scrutiny. Five questions to ask any provider:

1. What is the observed behavior, specifically? "In-market for luxury travel" is a label, not a methodology. Ask what event triggers inclusion. If the answer is modeled rather than observed, price it accordingly.
2. What is the refresh cadence? A segment refreshed monthly cannot support declared-intent activation. Ask for the actual recency distribution, not the maximum lookback.
3. What is the match rate against your identity graph? A segment with 4 million profiles and a 22% addressable match rate is a 880,000-profile segment. Plan against the real number.
4. What is the wealth verification source? Modeled income from ZIP-level census data is a weak proxy. Deed records, investable-asset panels, and verified financial relationships are meaningfully stronger.
5. Will they support a holdout test? Any partner unwilling to let you measure incremental lift against a matched control is telling you something about their confidence in the data.

Common Mistakes in Luxury Intent Signal Targeting

Stacking too many signals. Layering six intent conditions produces an audience of 40,000 people that costs $85 CPM to reach and cannot deliver meaningful frequency. Precision has a floor below which you cannot buy media efficiently. We generally target activation pools of 250,000–1.5M for national luxury campaigns, tightened through bidding rather than through audience exclusion.

Mistaking category interest for purchase intent. Someone reading three articles about private aviation may be a prospective buyer or an aviation enthusiast with no capacity to purchase. Intent signals must always be intersected with a wealth qualifier, not used alone.

Ignoring the household unit. For purchases above roughly $100,000, the decision is rarely individual. Targeting only the researching individual misses the spouse or partner whose approval is often decisive. Household-level identity resolution matters more in luxury than in any other category.

Failing to suppress. Closed-lost and closed-won records should suppress or redirect spend. Continuing to serve acquisition creative to a customer who purchased last month is both wasteful and damaging to the relationship.

Measuring intent audiences on last click. Intent-targeted media accelerates and concentrates demand, which frequently shows up as improved close rates and shortened sales cycles rather than more clicks. If your measurement cannot see cycle-time compression, it will under-credit the entire program.

What Good Looks Like

A mature luxury intent signal targeting program has four properties: a composite score back-tested against real closed-won data, decay logic tuned per signal class, first-party signals carrying the majority of activation weight, and incremental measurement through matched holdouts rather than platform attribution.

Brands that get there typically see qualified inquiry costs fall 30–50% within two quarters, not because they found a better audience vendor, but because they stopped paying premium CPMs to reach wealthy people who were never going to buy this quarter.

If you want your intent signal architecture audited against your own conversion data, apply to work with Stillwater Media. We take a limited number of engagements each quarter, and audience engineering is where most of the value gets created.

About Stillwater Media

Stillwater Media is a selective performance media agency for luxury and high-consideration brands. Based in Charlotte, NC and operating nationally, we build premium CTV, programmatic, DOOH, streaming audio, and affluent audience engineering programs for brands including JetLinx, W Hotels, PXG, FLY Exclusive, and Financial Independence Group. We accept a limited number of engagements per quarter. Signal. Strategy. Scale.