This has been a great discussion to follow. There’s been some thoughtful feedback mixed with a few misconceptions, so I figured I’d jump in and help clarify what xVIN actually is and how it works in real dealer campaigns. As I read through this thread and it’s been interesting to see how people are interpreting xVIN. A few of the assumptions being made come from older retargeting and DCO models that look similar on the surface, but the mechanics under the hood are very different.
xVIN isn’t DCO or cookie-based retargeting. The old “VDP retargeting” play followed a visitor around the web with cookies tied to a single VIN. xVIN doesn’t use cookies at all, and it doesn’t rely on prior site visits. It’s built on a custom DSP and DMP designed specifically for automotive, connected to a privacy-compliant identity graph that matches verified in-market shoppers to VIN-level inventory data across devices and channels.
Because of that, most of the people reached by xVIN have never even been to the dealer’s website. The audiences are built from verified shoppers who are researching vehicles across automotive marketplaces and same-brand or cross-brand dealer sites in the same market. These are real shoppers comparing models, trims, and pricing—precision prospecting powered by authenticated behavioral and intent signals, not site retargeting.
The part that usually surprises dealers is what happens next. Through a feature called InAudience, xVIN can connect campaign exposure to real retail outcomes by matching customer records from the dealer’s DMS—names, addresses, and sold VINs—to the identity graph. It’s not theoretical attribution; it shows which consumers were influenced, how they engaged on the dealer’s site, and which vehicles they actually purchased. While no automotive attribution model is ever perfect, no other VIN-level marketing platform provides this degree of visibility into real sales behavior. InAudience gives dealers a clear, data-driven view of how xVIN contributes to actual transactions—something cookie-based or third-party DCO systems simply can’t replicate.
There’s also an optional extension called DeepSignal that lets a dealer take the same verified xVIN audience into Connected TV. It delivers a full :30 spot—dealer branding, Why-Buy messaging, and dynamically inserted vehicles that match each viewer’s current interests. Traditional CTV can target by geography, demographics, income bands, context, and 1P/3P behavioral segments (and some providers use device graphs for household-level reach). What it generally doesn’t do is deterministically reach the same identified shoppers you’ve already engaged via xVIN based on their very recent VIN-level shopping signals—and then personalize the creative with specific matching inventory. That’s the gap DeepSignal fills, typically at comparable pricing to standard CTV.
Finally, the ads themselves aren’t just placeholders for VIN photos. Each is dynamically assembled from live inventory so the shopper sees vehicles that actually fit their current consideration set—make, model, body style, price range, and location—with current pricing and availability.
That distinction matters. A lot of “dynamic” ads in automotive—like typical social carousel or Automotive Inventory Ads—or even non-Meta platforms pull live inventory, but the targeting logic is broad and largely inferred or simply retargeted from the dealer's website visitors. xVIN takes a different approach: its personalization is powered by verified identity data and real-time behavioral signals, so the right vehicles reach the right shopper with VIN-level merchandising without the need of the shopper to first visit the dealer website. This is clearly not the “show them what they just looked at” logic from 2009.
That’s the real picture. If anyone wants to understand how it actually performs in live dealer accounts, I’m happy to walk through it—no sales pitch, just information. It may not answer every question, but it’s a lot better than guessing from the sidelines.
