AI Chatbots in Dealerships: What’s Actually Working in 2026?

Most dealership chatbots failed for one simple reason:

They were built to capture leads — not help customers.

In this episode of Dealer Tech Talk, Simon Verona revisits AI chatbot technology and explores how modern AI agents are evolving far beyond the old scripted website chat widgets dealers became frustrated with.

Today’s AI systems are becoming more conversational, more connected, and far more capable of supporting real dealership workflows.

But there’s a catch…

The success of AI in dealerships depends far less on the chatbot itself — and far more on how well it integrates with your CRM, DMS, processes, and customer journey.

In this episode, we discuss:

•  Why many dealership chatbots originally failed 
•  The biggest mistakes dealers still make with AI 
•  How conversational AI has rapidly evolved 
•  Why integration matters more than automation 
•  Where AI agents outperform humans 
•  Where human teams still massively outperform AI 
•  The future of AI assistants across chat, WhatsApp, and voice 

If you’re considering AI for your dealership — or wondering whether chatbots are finally becoming useful — this episode is essential listening.

👉 Sponsored by DMS Navigator & Miles Companion

Transcript

SPEAKER_00 0:00

A chatbot isn't your salesperson, it's a sales assistant. But the truth is that the problem isn't really AI in itself. It's the wrong AI implemented badly. We notice even that customers are forgetting that they're talking to an AI agent, often thanking them, wishing them a good weekend, for example. But of course, the AI doesn't have weekends. The chatbots are not dead, they're evolving. The basic ones are though. The dealers winning today are using smarter AI with a clear purpose. Welcome to Dealer Tech Talk, where we explore how technology is transforming car and motorhome dealerships. I'm Simon Verona. This podcast is sponsored by DMS Navigator, helping dealers become more efficient, deliver better customer service, boosting profits, and Mild Companion, the customer app that builds loyalty and keeps customers connected. Today's episode, revisiting chatbot technology in the dealership. A few months ago, I did an episode on AI chatbots in dealerships. And at the time, the promise was simple. Capture more leads, respond instantly, and never miss an inquiry. On paper, it sounds like a no-brainer. And it seems every day though, I see another AI chatbot launched for dealer use, all promising to be the best there is. And I've spoken to a lot of dealers who've tried these chatbots and not found them effective. So the question is, are the chatbots failing? Are they just being used in the wrong way? Let's be honest, most of these chatbot implementations don't deliver on their promises. And here's why. Firstly, the chatbot is asking too many questions. Customers are landing on your website, they click chat, and suddenly they're in a long interrogation that kills momentum instantly. The chatbot is being programmed to follow a script and gain a lead as quick as possible. The second one is that the responses are feeling robotic, they're not feeling human, they're feeling scripted, generic, and disconnected from what the customer actually wanted. And then the third one is that the chatbots often don't allow a human handoff. The customers get stuck with no easy way to get that onto a full onto a real person. And the fourth problem is really the key problem, and that's the lack of data, the lack of information. A lot of these chatbots are generic, they're preloaded with very little dealer-specific information and are therefore giving generic information back to the customer. And behind that, there's integration. Leads and that are coming through the chatbots aren't flowing into the CRM or your DMS system. And Sounds Exic can't always see the previous conversation in detail that the customers had with the chatbot. And the result is the dealers are not getting better leads, the customers are getting frustrated, and the chatbots are sometimes labelled as being not worth it. But the truth is that the problem isn't really AI in itself, it's the wrong AI implemented badly. And that's where things get interesting. Rolling forward a few months from when I last looked at it, the technology has moved on really quickly. And not all the AI systems you can see in the marketplace are equal. There are a good deal of relatively generic systems out there. But we're seeing a move now from the basic AI bots to ones that have full AI, data-driven, connected conversations, where the chat is more natural, less pushy, and the conversation is flowing more naturally in a human way. AI can now understand better customer intent. It can have a different conversation depending on where the customer is in the purchasing process. It can be a genuine assistant to the customer, helping them to make decisions, delivering useful information at the right time. And it can adapt in real time as the customer moves closer to making a decision. And behind that, they've access to live data and it can understand and consume that data better. So for example, if a customer asks a question, such as, Does the car have Apple CarPlay? Rather than just answer no or yes, it can modify that answer and suggest other similar cars in stock that do have that feature. And it can even go further than that and answer much more complex questions such as, will my golf clubs fit in the boots? But the key difference is now a far better integration with other dealer systems and processes. And this is important for logging conversations, booking appointments automatically to diaries, and handing over those inquiries to human sales executives. And more importantly, we're therefore moving beyond those simple chatbots. We're entering the world of AI assistants and AI agents. Instead of just answering questions, they'll be able to help move the deal forward over multiple channels. So a chatbot, which is handling customer inquiries on the website via WhatsApp, may later be able to hand over to a second AI agent, which will use the previous conversation to follow it up over a period of days and weeks, moving the customer from a top-of-funnel tire kicker into one which is indicate indicating intent and is ready to make an appointment to see a car and talk to a human sales executive. I see this in my day job. As you know, I write computer software for car dealerships. And we use AI technology extensively to deliver high levels of customer support. So very, very similar angle to what we've seen with car dealer chatbots. And we've seen a transition from simple question-answer sessions to deeper, richer conversational interaction between the AI agent and the human at the other end. Their interactions are feeling more human, and we notice even that customers are forgetting that they're talking to an AI agent, often thanking them to their better and wishing them a good weekend, for example. But of course, the AI doesn't have weekends. It's working 24-7, of course. So then when do chatbots work and how do they work? The key item that still jumps out is out-of-hours inquiries. In car sales, many inquiries are made outside of normal business hours. At the same time, many of these inquiries are quite early in the buying process, where the customer is qualifying which car to purchase and from what dealership. A good AI agent can handle these inquiries quickly and often better than a human in the same situation. But even during the day, when maybe a human could be available, the AI agent may well be more appropriate than a human to deal with inbound inquiries. It's quicker, it's more patient, and often has access to more information quickly than a human. AI agents are becoming particularly good at structured interactions with customers, such as booking appointments to see or test drive a car. This requires high levels of integration with the dealer diaries, the stock, and their CRM systems to complete the job. AI can also help in areas where compliance is critical, such as finance pre-qualification. An AI agent is able to apply the rules when discussing finance very specifically and help a customer with finance questions at the same time. The key item though is still answering those basic questions. Is a car still available? Does it have Apple CarPlay? The difference now is that it can answer those questions and provide information in a far more conversational way than the early AI chatbots. I would argue though that these sorts of use cases are often areas where your human team is not always the best and not the best use of their time. And it's important for the AI agent to see when the customer is moving beyond this into the realms of complex negotiations or even the complaint handling requirements or more high emotional conversations where it isn't the best tool for the job today, and it needs to hand that over to a human. And this is where the humans shine. One thing I noticed as our AI agent improved is that we have needed to add extra training to our team. Because they don't need to deal so much with the more basic tasks, they therefore need more training on the more complex items so they can handle and focus on them better. And what do I see the biggest mistakes that dealers are making today? The first is still perceiving an AI agent as replacing people instead of supporting them. A chatbot isn't your salesperson, it's a sales assistant. Its job is to capture, to qualify, to route, and then hand over to a human. The second is implementing an AI agent in isolation. This will never be anywhere near as effective as you'd hope. The agent needs to be implemented in a way that's integrated, both with your business processes and integrated with your other dealer systems and databases. So if you're thinking about using a chatbot, here's a quick and simple framework. Look at your work processes first and see where an agent can best be used to improve them. Website inquiries is an obvious item, but it isn't the only one. Think about the complete customer journey that the AI agent is involved in. Think about how it'll access information it needs, and how and when it will hand over to a human if necessary, and then integrate it properly with your systems and processes. But that's not a one-stop thing. You must continuously monitor and improve that. It isn't set and forget. Looking ahead, the space and the technology is evolving quickly. We'll see AI agents handling full conversations. What you now see as text chat will move to include voice. There'll be more personalization as the AI agent has access to customer and dealer data. And integration with dealer systems will improve more and more. So it isn't going away. The AI agents are just getting smarter. And here's the bottom line: the chatbots are not dead, they're evolving. The basic ones are though. The dealers winning today are using smarter AI with a clear purpose. Not replacing people, but supporting them and integrating with their existing workflows. Thanks for listening to Dealer Tech Talk. Don't forget to subscribe so you never miss an episode. This episode's been sponsored by DMS Navigator, my old companion. I'm Simon Verona, and I'll see you next time.