What’s next for automotive retail? 4 of the biggest takeaways from our expert panel

CEO Johan Sundstrand and Sophie Birshan, Google UK's Industry Head of Automotive, recently took part in a panel discussion at AM Live.

In a wide-ranging 45-minute chat, they covered everything from the marketing challenges facing dealerships to the practical applications of AI. Here’s four of the biggest takeaways to take into 2025 and well beyond…

Buyers are moving from showroom to search bar

Remember when selling cars was done mainly at the showroom? 

The reality is those days are long gone, and buyer habits have changed, with Sophie suggesting that many buyers now begin their journey online and ‘search without a brand in mind’. Instead, they begin their journey by searching for cars with certain characteristics, say best SUV for three kids and a dog with off-street parking.

What’s more, they aren’t just using Google but up to 7 different online-only sources when researching, including:

• OEM websites

• Comparison websites

• Review websites

• YouTube (which is growing 25% year-on-year)

• Dealer websites

And as Johan added, social media is something you also neglect at your peril. Over 80% of time online is spent on social channels consuming video, so if you’re not yet creating video content then you could be missing out on an awful lot of buyers.

Personalization is no longer a luxury; it's a necessity

Sophie suggested that in 2021, a buyer would typically compare one car brand with four others.

Fast forward to the present day and that figure has doubled. In fact, certain brands are compared to more than 10. 

You’ve got to get the right message to the right customer at the right time. This requires you to create more tailored content, and it’s here that AI can help. It’s now possible to create stills, inventory videos and paid ads in no more than a few clicks, meaning you can easily create content - at scale - for different demographics.

Johan gave the audience a walk-through of how simple this process can be, and if you’ll spare us a minute to use ourselves as an example, here's how Phyron makes this possible.

But AI isn't the answer to all your problems (...yet)

While it might feel counterintuitive to say it as an AI company, it’s the reality. 

A staggering 80% of all AI projects will fail, according to a study referenced by Johan, thanks to:

• Lack of quality data

• Outdated infrastructure

• AI tools being applied incorrectly

AI is often only as good as what you feed it, and to get the best results you need accurate, well-structured data to keep algorithms happy. Your IT infrastructure is important too - vintage tech just won’t cut it - and you need to set realistic goals for your AI projects. Testing and learning can help here, as Johan's comments on 'safe playgrounds' reflected.

As Sophie pointed out, it’s about starting small but thinking big. Pick one or two areas where AI can have an immediate business impact - say marketing at first - while you rectify any issues you might have on the data or infrastructure front. 

Once they’re fixed, AI will be in a better place to fix your more complex problems.

A retail revolution is coming in the form of AI agents

Right now, we’re using AI more as an assistant and as an extra brain.

According to Johan, this is just the beginning, and we’re moving to a place where AI will start acting as a ‘brain with legs’. It would make plans, execute on them, and require no further input from us, acting autonomously in our interests.

What this means for automotive retail is a world where autonomous agents interact with each other, with seller agents on one side and buyer agents on the other. This, Johan predicted, would make the process of buying a car ‘much quicker’.

Time will tell, but it’s an interesting question to ponder - what could buying and selling a car look like in the not-so-distant future?

If you want to watch the full panel, you can find it here. Grab a coffee, get settled, and follow us on LinkedIn for more content like this.