AI Lead Workflow for Car Dealers

For Episode 25 (December 19, 2025), Jordan vibe coded a workflow builder for car dealers after meeting the Auto Acquire team earlier that week. A web form fires a webhook, AI scores the lead's intent, and the flow either texts the customer or drops them into an email drip. He called it "mostly front end." He didn't name his tools or give a build time, so this page is on the short side.

Why we built it

In our pre-call, Auto Acquire's Anthony and Chris said a messaging bus would be useful for dealers. Jordan built a simple version of one. The audience shaped the design: dealers are mostly non-technical, so every step had to be readable at a glance.

The stack

Not named on the show. What the tool contains:

• A webhook trigger connected to a dealer's web form.
• An AI step that analyzes intent and assigns a confidence score.
• SMS and email branches with editable templates.
• A visual flow editor where each step can be changed, plus a button to run the flow end to end.

How we built it, step by step

1. Trigger on a web form. Someone fills out the dealer's form, which fires the webhook.
2. Analyze intent with AI. What does this person actually want, based on what they wrote?
3. Score it. The confidence threshold is editable, for example 90.
4. Branch. High confidence sends a templated SMS and continues the conversation. Low confidence goes to an email drip campaign for people who may not be ready to buy.
5. Run it. Execute the flow to test the whole path.

If we have a high confidence score, then we'll actually score it and then send them an SMS, like, hey, you know, why don't you come on by or something like that based on a template type system.

— Jordan Metzner, Episode 25

How it turned out

The live run at 03:09 worked. The test lead asked, "Is the 2021 F-150 still available?" and got an automated reply: yes, it's on the lot, would you like to schedule a test drive? Jordan admitted the CSS was "a little off."

Auto Acquire CEO Anthony Monteiro liked it. He said the days of big developer teams and slow review cycles are passing, though humans still need to prompt, test and connect the generated code to everything else. His suggestion was to flip the flow toward acquisition: ask an owner if their 2021 Porsche Cayenne is still available and whether they'd sell it for cash today, "and it doesn't tie up staff."

What we'd do differently

But what I like about the interface is I think it's it's very easily understood by a non technical user.

— Jordan Metzner, Episode 25

That's the part to keep. The next version would:

• Keep scoring the conversation, not just the form, and decide whether to keep texting, call, or bridge a call where the dealer's phone rings with the lead's details first.
• A/B test templates so dealers can swap out ones that don't convert.
• Adapt by dealer. Jordan guessed a small-town dealer's customers may prefer a call and a big-city dealer's a text. Anthony added that it depends on staffing: a dealer with a big BDC can hand off calls, others may need a voice AI agent.

More on AI in automotive. Sam co-hosted; he was selling two cars that week.

FAQ

How does AI lead scoring work for car dealers?

In Jordan's prototype, AI reads what the person wrote in the dealer's web form, judges their intent and assigns a confidence score. Leads above a threshold you set (for example 90) get an automated SMS; lower-confidence leads go to an email drip.

Can you automate SMS replies to dealership leads?

Jordan's demo did: a lead asking 'Is the 2021 F-150 still available?' got an automated text saying it was on the lot and offering a test drive. Replies come from editable templates.

What did Auto Acquire's CEO suggest?

Anthony Monteiro suggested flipping the same flow to acquisition, texting owners to ask if they'd sell their car with a cash offer, so the conversation runs without tying up staff.

What tools was the dealer workflow built with?

Jordan described it as mostly front-end vibe coding but didn't name the tools on the show.