What Is an AI BDC? How It Works, What It Costs and How to Choose One

An AI BDC answers and follows up on leads around the clock, but the stores that win with it keep people on money, credit and phone calls. Here is what it automates, what it costs, and how to evaluate one.

The short answer

An AI BDC is software that does the repetitive work of a dealership business development center: it answers new leads by text and email within seconds, follows up for months, books showroom appointments and sends reminders. A good one reads the lead source and the whole conversation, hands price, credit and "I want a real person" moments to your staff, and logs everything in your CRM. It replaces the speed and persistence part of the BDC job, not the judgment part.

What a dealership BDC does

A business development center (BDC) is the team between your leads and your showroom. Its job is to turn inquiries into appointments that show:

  • Answer new leads from third-party sites, the website, chat and phone-ups.
  • Follow up on everyone who did not buy, including missed appointments.
  • Set and confirm appointments for a specific day and time.
  • Work the database: lease maturities, service customers, past buyers.
  • Log it all in the CRM so managers can see what happened.

The work is measured as a funnel. In Foureyes' data on 2.6 million leads at more than 1,150 dealerships in the first quarter of 2026, the average store made contact with 69% of its leads, set appointments with 40% of the leads it reached, and saw 59% of those appointments show (Foureyes).

The weak spots are predictable: nights and Sundays, follow-up that fades when the desk gets busy, and turnover. Pied Piper's 2026 study of 3,290 dealership websites found that 51% of dealers gave a web lead a "perfect response" (fast, and through more than one channel), twice the rate of five years earlier (Pied Piper). Real progress, and roughly half the stores still fell short.

What an AI BDC actually automates

An AI BDC is software, usually built on a large language model, that holds real conversations by text and email (and, with some vendors, by phone). It reads what the customer wrote, answers it, and steers toward the next step. It should own four jobs.

First response

The first reply goes out in seconds at any hour, mentions the car or trade the customer asked about, and asks one question that moves toward a visit. In Pied Piper's 2026 study, dealers answered the customer's question by text 54% of the time, up from 38% a year earlier. The lead response time guide covers why the first minutes matter.

Follow-up

It keeps working every lead that did not book, for months or years, with a new reason to reply each time instead of "just checking in", and it stops the moment a customer opts out. Persistence is exactly what a busy human BDC drops first. For aged leads, see dead lead reactivation.

Appointment setting

It offers specific times, handles reschedules, writes the appointment to the CRM and alerts the salesperson. The goal is a day and a time, not "stop by this weekend".

Confirmations and reminders

It confirms right after booking, reminds the day before and the morning of, and reaches out within minutes after a miss. The show rate guide has the full cadence and sample texts.

What an AI BDC should not do alone

The worst failure of an AI BDC is not a clumsy text. It is a customer who needed a person and never got one. In Pied Piper's 2026 study, evaluations that required "Human Help" scored nine points lower on average, and those customers were twice as likely to get no personal response at all.

The greatest risk is in the handoffs; between systems and from AI to dealership staff. Teams assume automation is working and that someone will step in if it fails. Too often, no one does.Cameron O'Hagan, Pied Piper, February 2026

Put the handoff rules in writing before launch. These go to a person, fast:

  • Price, payment and trade value. A number quoted before anyone has seen the credit or the trade is a number you will have to walk back.
  • Credit and financing. No approvals, no rates, no "everyone gets approved".
  • Anyone who asks for a person, a manager or a callback. That request should reach a human's phone within minutes.
  • Complaints, legal threats and deals gone wrong.
  • Holds, deposits and special programs such as employee, military or first-responder pricing.

The rule of thumb: the AI owns speed and persistence, and people own judgment and money. Every escalation needs a named owner, a time limit and an alert that reaches a phone, and the AI should stay quiet until the conversation is handed back.

AI BDC vs human, outsourced and hybrid BDC

Most stores are choosing among four models. "Virtual BDC" is used loosely: some vendors mean a remote team of people, others mean software, so ask which one you are buying.

How the four BDC models compare
FactorIn-house BDCOutsourced BDCAI BDCHybrid (AI plus your people)
First responseFast when staffedDepends on the hours you buySeconds, any hourSeconds by text
Nights and weekendsOnly if staffedThe hours you buyCoveredCovered
Follow-up persistenceFades when busyThe vendor's cadenceMonths or yearsMonths or years
Phone callsYesYesOnly with an AI voice productYour people
Price and creditYour managersHand off to your deskShould hand offYour managers
Knows your storeBestFrom your trainingOnly what it is connected toAI for data, people for the rest
Cost structureWages, commission, benefits, a managerQuoted per storeSoftware fee plus messagingSoftware fee plus a smaller team
Biggest riskTurnoverReps who do not know your storeMissed handoffsUnclear ownership

For most franchise stores, hybrid is the realistic end state: AI answers instantly and never stops following up, and your people take the calls, the numbers and the showroom. Outsourced BDC firms are arriving at the same idea; Strolid describes itself as "Powered by People, Strengthened by AI" (Strolid).

What an AI BDC costs (as of September 2026)

Most vendors do not publish prices. Here is what was public in September 2026. Treat any price on a competitor's blog as a claim until the vendor puts it in writing.

Published pricing, checked September 2026
OptionPublished priceWhat to know
One in-house BDC repAbout $18.38 an hour on average (Indeed, updated September 20, 2026)About $38,200 a year at 2,080 hours, before commission, taxes, benefits, turnover and management
TECOBI$1,496 to $3,995 a month, plus a one-time setup fee of $499 to $3,995 (pricing)Varies by store size and term (month-to-month up to 12 months); SMS allowance with per-message overage
DealerAI$599, $1,499 or $2,299 per rooftop a month (pricing)Month-to-month with 30 days' notice; voice usage extra on the top plan; currency not stated
PodiumNot publishedAutomotive plans say "Get a quote" (plans)
DriveCentricNot publishedPriced by rooftop and units sold, on an annual or multi-year agreement (pricing page)
MatadorNot publishedFees set in each order form; auto-renews unless you give written notice 30 days before the term ends (terms)
NumaNot publishedNo prices on its pricing page, only a demo request
Outsourced BDC firmsNot publishedNo rates found on the sites of Strolid, Better Car People or Traver Connect
DealFloNot publishedMonth-to-month with no minimum commitment under its standard dealer agreement

Monthly price is the wrong comparison. Compare cost per shown appointment: everything you pay in a month (software, setup spread over the term, overages, and the people who work the handoffs) divided by the appointments the system set that actually showed. Run the same math on your current BDC.

AI BDC vendor evaluation checklist

Score every vendor on the same seven areas, and count a yes only when you see proof in the demo or the contract.

1. Lead-source awareness

  • Does it read the ADF lead source and open differently for a KBB Instant Cash Offer, a CarGurus listing, a website price request and a dead lead? See the ADF guide and KBB lead follow-up.
  • Does it read the customer's comments, not just the name?

2. CRM integration and write-back

  • Do leads arrive by ADF email, an API, or both?
  • What writes back (texts, call notes, appointments, status), and does it land on the existing lead record? See VinSolutions and eLead for CRM specifics.
  • How are duplicates handled when one shopper comes in from two sites?

3. Escalation rules

  • Which situations trigger a handoff, and can your managers edit them?
  • Who gets alerted, on which device, and what happens after 10 minutes of silence?

4. Compliance

  • Sending hours. Federal rules bar telephone solicitations before 8 a.m. or after 9 p.m. in the customer's local time (47 CFR 64.1200). Check state rules too.
  • Opt-outs. Under the same rule, "stop", "quit", "end", "revoke", "opt out", "cancel" and "unsubscribe" revoke consent, so does other wording a reasonable person would read as a request to stop, and revocations must be honored within 10 business days. Test it with "please stop texting me".
  • Registration. Ask whose brand and campaign are registered for A2P 10DLC, the standard US carriers use to keep business texting on 10-digit numbers verified and consensual (Twilio).
  • AI voice. In February 2024 the FCC confirmed that AI-generated voices count as "artificial or prerecorded voice" under the TCPA, so those calls need the called party's prior express consent (FCC 24-17).
  • Recording and disclosure. Some states require every party's consent to record a confidential call, California among them (Penal Code 632). California also bars using a bot online to mislead someone about its artificial identity to push a sale, with clear disclosure as a defense (Business and Professions Code 17941). Ask what the AI says to "is this a bot?" and have your attorney approve it.

None of this is legal advice. Put each vendor's answers in front of your counsel.

5. Reporting

  • Median first-response time, by source and by hour.
  • Contact, set, show and sold rates, split by AI-set versus people-set appointments.
  • Time to a human on every escalation, plus opt-outs and complaints.
  • Full conversation transcripts, not just summaries.

6. Call handling

  • Does it answer calls with an AI voice, route them to people, or leave calls alone?
  • Does the next text know what was said on the last call, and are calls recorded and scored for coaching?

7. Contract terms

  • Initial term, auto-renewal and the notice window to cancel.
  • Setup fees, message allowances and overage rates.
  • Who owns the phone numbers and conversation data, and how you export both.

Questions to ask in an AI BDC demo

A demo is a sales call. Turn it into a test by asking the vendor to run these live:

  1. Submit one test lead as a KBB Instant Cash Offer and one as a CarGurus lead. Do the first texts differ, and does each ask for a specific time?
  2. Reply "what's your best price?" Who at your store gets alerted, and how?
  3. Reply "please stop texting me" without the word STOP. Does it stop?
  4. Ask "am I talking to a real person?" Would your attorney sign off on the answer?
  5. Ask about a car that sold yesterday. Does it know?
  6. Show us three unedited conversations from a store like ours, including one that went wrong.
  7. Show us exactly what lands in our CRM after this conversation, field by field.
  8. What should our show rate and cost per shown appointment be at 90 days, and will you put that in writing?

Watch who answers. If only the salesperson can explain how escalations work, ask to meet the person who will support your store after you sign.

How to measure an AI BDC at 30, 60 and 90 days

Before go-live, pull a 90-day baseline by lead source: response time, contact, set and show rates, units sold and opt-outs. Without it you will end up arguing about anecdotes.

What to check and when
CheckpointWhat to checkWhat good looks like
Day 30Every lead answered, nights and weekends included; escalations reached a person on time; CRM write-back matches a 20-lead spot-check; opt-outs honoredNo unanswered leads, no missed escalations, no compliance misses
Day 60Contact, set and show rates by source against baseline; 25 conversations read in full; reminders going out on scheduleRates above baseline on the sources the AI touches
Day 90Units sold from appointments the AI set or rescued; cost per shown appointment; aged leads that re-engagedCost per shown appointment at or below your current BDC

Compare against your own baseline, not a national average. In Foureyes' Q1 2026 data, show rates ran from 41% to 74% depending on the market, and appointment set rates from 30% to 59%.

Frequently asked questions

Is an AI BDC the same as a chatbot?

No. A website chatbot waits for visitors. An AI BDC works leads from every source, starts the conversation by text or email, follows up for months and books appointments into your CRM. Some products do both.

Can an AI BDC replace my BDC staff?

It can take over first responses, follow-up and reminders. It should not replace the people who handle calls, price and credit questions, and customers who ask for a person. A realistic setup is a hybrid.

How much does an AI BDC cost?

Most vendors quote privately. As of September 2026, TECOBI lists $1,496 to $3,995 a month plus a setup fee, and DealerAI lists $599 to $2,299 per rooftop a month. Compare vendors on cost per shown appointment, not monthly price.

What is a virtual BDC?

Usually an outsourced team of people who answer your leads and calls remotely under your store's name. Some AI vendors use the term for software, so ask whether you are buying people, AI or both.

Is AI texting legal for car dealers?

Texting customers is regulated, not banned. Federal TCPA rules cover consent, sending hours and opt-outs, and state laws such as California's bot disclosure law add more. Have your counsel review how each vendor handles consent, STOP replies and "is this a bot?"

How long does an AI BDC take to set up?

Receiving leads by ADF email can be quick, since it only means forwarding a copy of your lead feed. Writing conversations and appointments back into your CRM depends on the vendor's integration with your specific CRM, so get that timeline in writing.

Josh SkwarekFounder & CEO, DealFlo

Josh Skwarek is the founder and CEO of DealFlo, the lead-handling system for car dealerships that answers every lead in seconds, grades every sales call and follows up with every customer until they buy.

See it run on your own leads

A 30-minute walkthrough on your store's real lead sources: the first text, the graded call, and the follow-up that never quits.