The Auction Has Competition: One Million Agentic AI Conversations Turn Dealers' Own Databases Into America's Cheapest Car Source

inride's Trade Agent AI crosses one million customer text messages across 71 rooftops in 18 states — finding 65.4% of customers are sitting in positive equity, averaging $13,763, and driving cars that out-gross auction inventory by 73%.

BETHESDA, Md., Aug. 12, 2026 /PRNewswire/ -- American car dealers are paying record prices at auction for used inventory. According to one million Agentic AI-run customer text conversations, the better cars are in their own database.

Trade Agent AI, the vehicle acquisition agent built by inride.ai, has now handled more than 1.1 million text messages across nearly 225,000 customer conversations inside dealership databases and measured something the industry has never seen at this scale:

65.4% of a dealer's own past customers are in positive equity with their current vehicle, by an average of $13,763. 

Edmunds reports that 68.8% of trade-ins toward new-vehicle purchases carried positive equity in Q2 2026, at a record average of $13,330 and those are only the customers who already showed up at a dealership. Trade Agent AI measures the entire database, in-market or not, and finds nearly the same opportunity sitting in the files at an even higher average. Most of those vehicle owners have no idea. Until now, neither did their dealers.

The average dealership database holds tens of thousands of past customers, and finding the sellers among them requires an individual conversation with every record. This is volume no BDC on earth has the hours to run. Trade Agent AI runs it. The agent texts and emails every customer in a store's database, values their vehicle, surfaces the owners with real intent to sell or trade, and books hard appointments with a live date and time.

What one million messages measured:

• 41.4% initial response rate. More than four in ten of the nearly 225,000 customers contacted responded to the agent. (excludes opt-outs)
• 79,440 engaged conversations — 35.4% of everyone contacted. More than a third of the database held a real, two-way conversation with the agent about their vehicle.
• 11,623 converted to leads as owners expressed true intent to sell or trade — all identified and sent directly by Trade Agent AI to dealers' CRM with no human involvement.
• 11.3% of those warm leads became hard appointments set entirely by the Agentic AI — a confirmed date and time, booked end-to-end with no BDC, no call center, and no human involvement.
• 4.5 individualized touches per customer, at a scale no store could staff. Matching that volume by phone at the industry benchmark of 50 quality BDC contacts per agent per day would take nearly 78 full-time agents.

What the cars are worth when they come back:

• A 73.2% gross premium — with no buy fee, no transport, no stranger's car. Database-sourced vehicles retail at $3,176 average gross against the participating stores' used-vehicle average of $1,834 — $1,342 more per copy than the cars dealers fight for at auction — while every acquisition avoids a dealer-reported average of $1,500 in all-in auction cost, before counting the recon and days-to-frontline advantage of a vehicle the store originally sold and serviced.
• Two (or more) retail opportunities per vehicle, not one. A database trade arrives attached to a purchase, then retails again. Vehicles that have completed both legs did so in an average of 18 days from acquisition to resale, and each acquisition has generated $3,544 in total dealership gross to date — a number still climbing as inventory turns. 

One VIN tells the whole story. A Midwest store sold a customer a used 2022 Grand Cherokee L in March 2025, then had no way of knowing when she'd be ready to move again. Trade Agent AI reached her on June 16, 2026. On June 25 she bought a new 2026 Grand Cherokee L and traded the 2022 back in — a deal that grossed $6,590. Ten days later, the store retailed the 2022 to its next owner for another $9,598 in gross. Two transactions. $16,188. One customer the store already had, found by an agent that never sleeps. And the chain is still running: the 2022's new owner now sits in that same database as a future trade, a future retail deal. Every VIN acquired this way carries its origin story forward into business the store would never otherwise have seen.

Nick Perdikis, CEO of inride: "Dealers do not have an inventory problem, they have a conversation problem. The cars they need are already in their customers' driveways, and 65% of those owners are in positive equity and have no idea. We built Trade Agent AI because the only thing standing between a dealer and that inventory was a million customer messages nobody had the hours to handle. Now they have been handled."

Josh Barbour, General Manager of Capital Ford in Wilmington, N.C., a Trade Agent AI store: "We have found that through working our owners, using the power of AI, we are able to acquire the right cars, at the right money, from people that already want to do business with us. Additionally, we are able to continue the relationship with them by trading those customers into something new."

A note on methodology. Every figure in this release reflects only activity inride can directly trace. Attribution requires the agent to have contacted the customer before the deal closed, and inride does not claim the agent was the customer's only touchpoint. Attribution is ongoing — which means the rates and averages above are floors, not estimates.

Availability. Trade Agent AI is live on 71 rooftops and adds stores weekly. To learn more or schedule a demo, contact [email protected] 

About inride

inride.ai builds AI agents for automotive retail, founded from both sides of the problem it solves. Founder Alex Perdikis is a 30-year dealer operator and owner of the Koons Motors group in the Washington, DC market, where Trade Agent AI has been live on his own showroom floors since 2025. CEO Nick Perdikis brings more than 35 years in SaaS as a founder, owner, and operator to the product side, pairing dealer-grade instinct with enterprise-software discipline. inride's flagship, Trade Agent AI, runs customer conversations at scale to source used-vehicle inventory directly from dealer databases, replacing auction dependency with owner-to-dealer acquisition. inride is based in Maryland. More at inride.ai.

SOURCE inride