How automotive retailers use AI to improve sales and service

Automotive retailers use AI to answer sales and service enquiries at any hour, move customers towards a decision before a salesperson steps in, and give teams time back. Keyloop’s data suggests the gains come when AI helps the customer first and asks for details second. 

Most of the opportunity is lost at the first step. In a Keyloop mystery shop across 10 markets, including the UK, 55% of dealer and manufacturer leads got no response of any kind. And too many dealer chats still end with “a colleague will be in touch”, which leaves the customer in a queue while their intent fades. 

To see what works better, we analysed more than 4,000 real conversations with AIME, Keyloop’s AI agent, for the AIME Report 2026. This article uses that data to look at what customers are really asking, when the best conversations happen and how to tell whether AI is working. It then shows where AI fits in with effective vehicle sales and aftersales. 

The questions today’s car buyers are really asking 

AI helps most when it recognises the decision behind the question. Every customer question has two layers, the one they ask and the one they’re working through. 

These sound like information requests. The data suggests they’re closer to purchase decisions in progress. 62% of the conversations we analysed began with a customer already progressing a purchase decision, and 71% included the customer typing concrete data about themselves or their car. Customers volunteered budget, mileage, preferences, timing and vehicle requirements when the conversation helped them move forward. 

When should AI ask for car buyer contact details? 

Our data points to once the customer has been helped, not before. Conversations that ended in a lead ran to a median of 22 messages, against 7 for those that didn’t. Overall, 37% of conversations ended with contact details captured. 

Not every customer is ready to leave their details, and that’s normal. In the conversations that didn’t convert, 51% still returned information such as a vehicle, a price or a service answer. 34% were asked for contact details and chose not to leave them, and of those, 34% had already received what they asked for. 

Even a disappointing answer keeps the conversation going. 56% of customers name the figure they had in mind when prompted, and 35% of customers whose valuation disappointed them go on to book a callback to discuss it. By the time a salesperson picks up, 66% of leads already include a part exchange valuation and 93% contain more than a name and a phone number. 

When do the best conversations happen? 

Often when nobody is on the forecourt. 40% of the conversations we analysed happened outside working hours, and 45% of all qualified leads were captured then. They convert well too. Between 21:00 and 01:00, 46% of conversations became a qualified lead, compared with 32% between 09:00 and 13:00. Saturday was the strongest day, at 43%. (Keyloop, AIME Report 2026). 

Why one conversation is rarely enough 

Buying a vehicle isn’t linear. Customers explore, leave, compare and return. 21% of conversations were with returning customers, and 29% returned within six hours with a further 29% returning within 24 hours. 39% of customers who left without converting became a lead when they came back, and second conversations were shorter, averaging 6 messages instead of 12. 

How to tell whether AI is working 

Measure progress, not handoffs. Retailers have become experts at measuring milestones such as visits, leads and enquiries. But the strongest signals of intent often emerge between them, which changes what’s worth measuring. 

The goal isn’t to end conversations quickly. It’s to move customers closer to a decision. 

How AI improves service and aftersales enquiries 

The right automotive AI agent handles service as well as sales. It answers questions on MOTs, servicing and availability the moment someone asks, day or night. It then either directs customers to your service booking link or takes a request for your team to follow up. 

For the booking and scheduling side of aftersales, read our guide to reducing workshop no-shows.  

Why governance matters when AI talks to your customers 

Trust matters when AI talks to customers. In Cisco’s 2024 Consumer Privacy Survey, 59% of consumers said strong privacy laws make them more comfortable sharing information in AI applications. 

AIME’s responses are grounded in approved data, supported by automotive-specific workflows and protected by built-in guardrails that keep conversations accurate, relevant and auditable. Scope controls and human escalation are part of the design, and customer data is never used to train AI models. Read more in the Keyloop Trust Center. 

What retailers report 

Retailers using AIME report returns of £18 to £37 for every £1 invested. (Source: AIME customer results, 2026) Infinity Motors, an independent retailer, reports that AIME saves its sales team around 30% of their time. Two more share their numbers below. You can read more customer success stories, including Highfield Motors Group’s AIME story. 

Scott Kingsley, Manager at Junction Autopark, says: “Before AIME, we were getting chat leads that simply weren’t worth pursuing. AIME changed that completely.” Read the National Car Hub story too.