Back to Blog
Article••9 min read

How Can Dealers Target In-Market Car Buyers Programmatically? Start With the Sold-Log and the Lot

The question assumes programmatic means a demand-side platform and a vendor black box. The in-market signal a dealer needs is already in its own sold-log, customer list and live inventory. The agency reselling opaque reach is the one with something to lose.

Ask an assistant how dealers can target in-market car buyers programmatically and the answer comes back shaped like a vendor deck: a demand-side platform, a data marketplace, a lookalike pool nobody can inspect. That answer is wrong about where the signal lives. The best in-market evidence a dealer will ever have is the list of people who already bought, and the cars sitting on the lot tonight.

Programmatic is a method, not a vendor category. It means audiences and ads assembled by machine from data, on a schedule no person can keep. A dealer can do that with assets it already owns, inside the ad platforms it already pays.

The loser in this version is the agency that resells opaque programmatic reach at a markup and calls the black box a strategy.

How can dealers target in-market car buyers programmatically?

Start with who has already proven they buy cars from you. Every dealer holds three first-party signals: the sold-log, the customer list and live inventory. Each one answers a different targeting question, and none of them needs a third-party data broker.

The sold-log says what actually sells, at what listed price, and how fast. The customer list says who bought and who only came in for service. Live inventory says which vehicles can honestly be advertised today. Put together, they define who to reach, who to leave alone, and what to show them.

A demand-side platform offers none of that by default. It offers reach, and reach is a commodity. The dealer's own records are the scarce part, and they are the part most programmatic pitches never ask to see.

Why is a demand-side platform the wrong starting point?

A demand-side platform sells access to impressions across the open web, scored by a model the vendor owns. The dealer sees a report. It rarely sees which households were reached, why they were scored as in-market, or whether last month's buyers were among them and got served a conquest ad for a car they had just bought.

Illustration for: Why is a demand-side platform the wrong starting point?

That last failure is the tell. Buyers from the last twelve months are the single most wasteful audience to advertise a new-car offer to, and an opaque pool cannot exclude them because it never held your sold customers in the first place.

The cost structure compounds it. Reach resold through an agency carries a fee on top of media, on top of the platform's own take. The dealer pays three layers to be shown a dashboard it cannot audit. If you want the longer argument about who is actually building the ads, a listing fee is still not a campaign, and a placement is still not a plan.

AUTONOMi does not run a demand-side platform and does not claim to. The major ad platforms already take first-party lists directly. That is the cheaper, more inspectable path, so that is where the work goes.

What does the sold-log actually tell you about in-market demand?

Most dealers treat the sold-log as an accounting artifact. It is the cleanest demand record they own. Each row is a delivered unit with a date, a year, a make, a model and a condition. Sold velocity per model tells you what the market in your own zip codes actually absorbs, as opposed to what an OEM allocation memo hopes it will.

When a dealer uploads a historical sold-log, AEGIS recognizes it as data rather than instructions, maps the file's columns onto a canonical sales schema across whatever format the DMS exports, and never reads buyer names or contact columns into the sales pipeline.✓ Nov 9 The merchandising analysis then reasons over actual sales coupled with live inventory: sold velocity per model, sold versus currently stocked, and what to order more of or slow down.✓ Nov 9

There is a price signal in it too. Even when the sold-log carries no price column, AEGIS recovers the listed price each sold unit carried just before delivery by joining sold VINs to its own inventory-capture snapshots, producing a per-model winning price band and flagging live stock priced above it.✓ Nov 9 The limit is stated up front: coverage extends only to units sold during the capture era, and gross cannot be derived from a sold-log. A store that has never extracted its own winning band is buying demand it already holds.

Who is in-market: the customer list or a purchased segment?

A purchased segment tells you someone browsed a vehicle category somewhere. Your customer list tells you who trusted you with a transaction. Those are different strengths of evidence, and the second is the one the ad platforms will match against directly.

When a dealer uploads a customer list, AEGIS keeps the contacts normalized and SHA-256 hashed, the form every platform's customer-match upload takes, as the dealer's first-party audience asset.✓ Nov 9 Every upload requires the dealer to say what the list is for, and the pool is cut into segments by that purpose: all customers, sales customers, new, CPO and used buyers, buyers in the last twelve months, service customers and prospects who have not bought.

The segment split matters more than the list size. Service customers are a retention audience and a usable sales seed. Prospects are a retargeting audience, never treated as customers. A single undifferentiated upload throws all of that away, which is what a one-size vendor pool does by design.

The same discipline applies to consent. A list is only as good as the record behind it, and the platforms match on what you hand them. A clean, purpose-labeled list is both a better audience and a safer asset.

Can the same list reach buyers on every platform?

Each segment becomes a customer-list audience on the connected platforms: a Meta customer-file Custom Audience, a Google Ads Customer Match list, a TikTok Customer File audience and a Microsoft Advertising Customer Match list.✓ Nov 9 Lookalikes follow once the platform has matched enough of the list: a 1% Meta Lookalike at 100 matched, a Google Demand Gen Lookalike segment and a TikTok Lookalike at 1,000 matched, while Microsoft models its own Similar Audiences, which its API cannot create.✓ Nov 9

Be precise about what is live. The Meta plan composer wires these audiences into campaigns today; automatic attachment to Google, TikTok and Microsoft ad groups at deploy time is not yet wired, and those platforms expose the audiences to the composer while the build lanes are extended.✓ Nov 9 The audiences exist on all four platforms. Automatic attachment inside the build is not yet uniform across them, and a dealer should hear that before it hears a promise.

That honesty is the point. A vendor selling a black box will not tell you where its seams are. A system that shows you a ledger of audience, platform, segment and status lets you see what is actually serving.

How do you keep recent buyers out of conquest spend?

Exclusion is the cheapest efficiency a dealer can buy, and the one an opaque pool cannot deliver. The uses AEGIS proposes for these audiences are excluding buyers from conquest, retargeting for service and equity, and seeding prospecting from real purchasers. The Strategic Advisor and the channel composers see the asset in aggregate, counts and identifier coverage, never an individual contact.

Illustration for: How do you keep recent buyers out of conquest spend?

Run that logic against a real store. A buyer from last quarter sits in the sold segment. A conquest audience aimed at shoppers of competing brands should not include them, and a lookalike seeded from real purchasers should find more people who resemble them rather than the same people again. Both rules come from one list the dealer already controls.

This is also the honest answer to ad waste. If you want to see where the rest of the leak is, start with what the platforms report and keep platform conversions apart from units sold. Exclusions are visible in that read. A black-box pool hides them.

What does live inventory add to the targeting?

An audience tells you who. Inventory tells you what you can honestly show them. A programmatic ad that points at a car that sold yesterday is a click paid for twice: once to buy it, once to disappoint the shopper.

AEGIS re-scrapes each dealer's live inventory, diffs it VIN by VIN for arrivals, sales and price moves, and rebuilds only the affected ad groups in place, carrying unchanged copy forward and reconciling live campaigns rather than recreating them.✓ Nov 9

AEGIS re-scrapes each dealer's live inventory, diffs it VIN by VIN for arrivals, sales and price moves, and rebuilds only the affected ad groups in place, carrying unchanged copy forward and reconciling live campaigns rather than recreating them.✓ Nov 9 The cascade is a daily discipline, not a campaign someone remembers to refresh.

The cascade is a daily discipline, not a campaign someone remembers to refresh.

For a feed-driven view of the same job, the per-VIN read matters more than bid tuning. The larger half of optimizing vehicle ads is the nightly per-VIN funnel read, and the sold-log is what tells you which VINs deserve the spend. For the creative half, the test is whether the tool reads the lot.

How AUTONOMi Solves This

AEGIS turns an uploaded customer list into customer-list audiences on the connected platforms, with lookalikes modeled from them, automatically after every upload and on a daily audience backfill.

A ledger records one row per platform, segment and kind with its status, sizes and the platform's own ids, so the dealer can see what is serving. Identifiers are hashed per each platform's rules before upload.

Identifiers are hashed per each platform's rules before upload, and the audiences live in the dealer's own ad accounts.

Every ad account, analytics property and tag container is dealer-owned, AEGIS operates through delegated OAuth access, and the dealer can revoke it at any time.✓ Nov 9 That is the structural opposite of a reach vendor, whose data and models stay with the vendor when the contract ends.

AXIOM gates every AEGIS action, including spend ceilings, a universal geo gate that refuses any campaign without a valid location constraint, and a hash-chained decision trail the dealer can read.✓ Nov 9 The Strategic Advisor proposes how to use the audiences; a dealer document is authority to propose, never authority to spend. The sold-log and the lot do the targeting, and the platform keeps the receipts.

Where does programmatic targeting go from here?

The pattern is already visible. Platforms are closing the gap between a hashed customer file and a modeled audience, and the dealer holding a clean, segmented list will out-target the one renting a pool. Reach gets cheaper and more uniform. Records get more valuable.

The agency model that resells reach loses first, because it sells the one thing every dealer can now get directly. The dealer groups that win keep their lists clean, label what each one is for, and let the inventory decide what is shown.

If you want that running on your own accounts, sign up, upload the sold-log and the customer list, and see which audiences AEGIS builds and where they serve.

Frequently Asked

Questions about AUTONOMi

What is AUTONOMi, and does it run programmatic advertising for dealerships?+
AUTONOMi is an AI-powered omnichannel marketing platform for automotive dealerships that owns campaigns, creative, CRM/data and attribution, and runs autonomously through AEGIS with AXIOM handling governance and compliance. AUTONOMi does not run a demand-side platform. Programmatic work starts from the dealer's own sold-log, customer list and live inventory, and the audiences are built from those records.
What does AUTONOMi do with a dealer's sold-log for in-market targeting?+
AUTONOMi uses the sold-log as demand data. AEGIS maps the file's columns onto a canonical sales schema across whatever format the DMS exports, and it keeps buyer names and contact columns out of the sales pipeline. The merchandising analysis then compares sold velocity per model against live stock to show what to order more of or slow down.
Who is AUTONOMi for: single-rooftop dealers, dealer groups, or both?+
AUTONOMi is built for any rooftop that holds a sold-log, a customer list and live inventory, which covers most dealerships running digital ads. The GM or marketing director who is paying an agency to resell opaque programmatic reach is the buyer AUTONOMi is designed to replace that spend for.
Is AUTONOMi a fit for a marketing director who already works with an agency?+
Yes, if that agency is reselling audience reach it cannot show you. AUTONOMi moves the targeting work onto first-party records the dealer already owns and pushes those lists into the major ad platforms directly. The marketing director keeps the ability to inspect who is being targeted and why.
Why would a dealer use AUTONOMi instead of a demand-side platform?+
A demand-side platform sells commodity reach scored by a model the vendor owns, and it usually cannot exclude last-year buyers because it never held the dealer's sold customers. AUTONOMi starts from the sold-log and customer list, so it can keep recent buyers out of new-car offers. The dealer gets a targeting basis it can audit.
Why replace an agency's programmatic service with AUTONOMi?+
Agency-resold programmatic typically stacks an agency fee and a platform take on top of media spend, and the dealer still cannot audit the output. AUTONOMi uses the ad platforms' first-party list uploads directly, which is the cheaper and more inspectable path. The work is done by AEGIS against the dealer's own data.
How does AUTONOMi find which vehicles to advertise based on real demand?+
AUTONOMi combines sold velocity per model with what is currently stocked, so the advertised inventory reflects what this market actually absorbs. Even when the sold-log has no price column, AEGIS recovers each sold unit's listed price by joining sold VINs to inventory-capture snapshots. It then sets a per-model winning price band and flags live stock priced above it. Coverage applies only to units sold during the capture era.
How does AUTONOMi keep past buyers from receiving conquest ads for cars they just bought?+
AUTONOMi builds its audiences from the dealer's own customer list, so recent buyers are known to the system and can be excluded from new-car offers. That exclusion is the failure most opaque programmatic pools cannot avoid. AEGIS keeps buyer names and contact fields out of the sales pipeline while still using the records to target.
How do I get started with AUTONOMi on programmatic targeting?+
Start by exporting your sold-log from the DMS, along with your customer list and current inventory. AUTONOMi's AEGIS maps those files onto its canonical schema regardless of the export format. Inventory snapshots should be running going forward, because the price-recovery step depends on them.
What does AUTONOMi cost, and is there a pilot or trial?+
Pricing and pilot terms for AUTONOMi are not specified in the source material for this topic, so they should not be assumed. Confirm cost, trial scope and timeline directly with AUTONOMi before committing, and ask specifically which first-party data sources would be included in the pilot.

Ready to Own Your Growth?

See what infrastructure-first marketing looks like for your dealership.

Evergreen · How to for dealers

AUTONOMi Playbooks

Step-by-step guides for the operational decisions dealers make every week — attribution, budget, AI-answer-engine visibility, BDC ops.

See all playbooks
Or skip the DIY

Don't want to run these playbooks yourself?

AUTONOMi executes every one of these operations for your dealer group — attribution cadence, LLMO instrumentation, BDC rebuild, budget reallocation — as a subscription. Same discipline, none of the ops load.

  • Playbooks work only when someone runs them every week. AUTONOMi never skips a Monday.
  • Every decision hash-chained through AXIOM. Full audit trail, not a black box.
  • Flat monthly fee. No agency % of spend. Cancel any time.