As AI moves faster than my neighbour’s dog when our cat strays into its garden, I’ve been considering what agentic AI could mean for car dealerships and the role Autofinity’s developers believe we can play, writes Andy Whitehair, chief executive of Autofinity.
Hang on, what is agentic AI?
Most of the AI currently in use, and with which we are becoming familiar, is generative. Generative AI writes text, creates content or answers questions, including through the customer service avatars that have become increasingly commonplace.
Agentic AI takes the technology into a different universe. It can reason, make decisions and complete a series of steps to see a task through from beginning to end. It doesn’t just talk the talk, it walks the walk.
Sounds great: lower costs and a 24/7 worker that never needs a holiday
Certainly, although some of the fears raised by Alan Milburn’s review of youth unemployment suddenly come into sharp focus. The review warned that the growing number of young people not in employment, education or training could be costing the UK around £125 billion a year.
Employment is not the only issue at stake
Handing decision-making powers to AI is a little like jumping into a driverless taxi. We don’t quite trust it yet, particularly when dealers are operating on tight margins. Who authorised that discount? Then there is the compliance question. Try blaming the algorithm when the FCA comes knocking.
Nevertheless, the transition from chatbot to digital employee is happening rapidly. It may not replace sales executives just yet, but imagine a customer asking a digital assistant to find a silver hybrid hatchback with fewer than 25,000 miles on the clock, within 30 miles of their home, and then book a test drive.
That sounds reasonable and rather clever
I agree. The customer’s digital assistant could communicate with AI agents working for manufacturers, dealers, independent businesses and online marketplaces, producing a shortlist of suitable vehicles from which the customer can choose.
It saves the car buyer from trawling through numerous websites to find something matching their requirements.
Do I just need better descriptions to ensure my cars are considered?
Partly, but it goes much deeper than having the right vehicle descriptions and images.
An AI agent does not necessarily have to extract information from a dealer’s website. Where authorised, it could query data held directly within dealership systems to identify the vehicles that best match the customer’s requirements, provided it has access through a secure and intelligent API.
What if the customer never visits my website? What happens to finance and value-added products?
As the technology develops, agentic AI applications may move beyond reading website information and begin accessing approved data held securely across the systems operating within a dealership.
An AI agent could negotiate on behalf of the customer, help assess a part-exchange and arrange a test drive. On the other side of the transaction, the dealer’s AI agent could produce a detailed proposal incorporating finance and value-added products.
Connecting the different stages of the vehicle purchase could speed up finance workflows while supporting consistent, auditable processes designed around FCA requirements.
That brings us back to the central question: how much authority are we prepared to hand over to AI?
Could AI take control of the entire journey, from choosing the car to completing the transaction?
This is one of the conundrums facing not only automotive retail but society more broadly. If we continually hand decisions to AI, there is a danger that our own decision-making skills will weaken.
In automotive retail, however, I cannot imagine many car buyers opting out of the entire process. A vehicle remains too important, expensive and emotive a purchase.
Customers are more likely to ask AI to complete the initial search and create a shortlist. The quality of those results will depend both on the instructions supplied by the customer and the standard of the data held within dealership systems.
One opportunity Autofinity is exploring is enabling AI to scan group inventory and identify vehicles that have remained in stock for too long. That could bring significant benefits for dealers.
Agentic AI could also flag operational problems, detect subtle shifts in the market and suggest an appropriate response, potentially keeping dealers one step ahead.
For customers, much of the traditional friction associated with buying a car could disappear. Emails, finance paperwork and some of the cut and thrust of the showroom negotiation may be replaced by machine-to-machine communication.
Does that mean I won’t need a sales team?
Dealers will still need people to provide the human touch, create an emotional connection and build relationships.
Buying a car will remain a major decision and customers will still want to feel valued, particularly during the vehicle handover. However, sales leads could become even more dependent on the quality of the dealer’s back-end data.
Whether it is generative or agentic, AI should ultimately improve productivity and give dealership employees more time to talk to customers.
Our dealership data needs to be in good health
It needs to be exemplary, accurate and trustworthy if AI assistants are going to present it to customers, interact with other systems and influence the choices that result in one vehicle being selected over another.
A genuinely connected ecosystem will need to emerge, with little room for walled gardens. Agentic AI will only be as effective as the data it can access and the boundaries that people place around its actions.
Siloed datasets remain a major problem within automotive retail. They could prevent machine-to-machine transactions from taking place.
The customer’s AI assistant may simply move on to another dealer. The first dealership may never know it was overlooked, despite having precisely the right vehicle in group stock.
Worse still, an overly enthusiastic dealer agent could make a costly mistake in its determination to complete the task.
Handing that much power to machines still makes me nervous
This is where we see Autofinity playing a further role.
We have established ViHUB as a central operational hub through which updates can populate integrated systems. Dealers can access a single view of an individual vehicle and their wider group inventory, with actions highlighted through the dashboard.
In a world of machine-to-machine customer experiences, dealerships will need effective guardianship. They must be confident that AI-led processes remain within their commercial parameters, support FCA compliance and do not allow automated negotiations to erode margins.
We see ViHUB providing the guardrails that keep AI assistants within the boundaries established by the dealership, protecting both dealers and customers.
We are at an exciting and pivotal moment as agentic AI begins to embed itself within the customer experience. Dealers need to be open minded about its potential while remaining alert to the risks and putting the correct safeguards in place.
ViHUB already allows dealers to establish tailored rules of engagement. As AI agents become more capable and autonomous, those rules will become increasingly important.
Author: Andy Whitehair, chief executive of Autofinity.
Ensure you always get AM insight. Make us a preferred source of news on Google.