AUTOMOTIVE: How AI Is Changing the Rules of Search

If your dealership website is seeing fewer visitors than it was a year ago, you’re not imagining it. Some of that declining traffic is a direct result of how artificial intelligence (AI) is reshaping the way people search online and the rules are changing faster than most dealers realise.


Over the last 18 months as a research project I’ve been studying the effect of AI technology on Internet search. New agency disciplines like Generative Engine Optimisation (GEO) Conversational Search (CAI) and Answer Optimisation (AO) all drive AI Search Summary results including the citation links.

The fact is – Customers are increasingly getting the answers they need in the AI Summary at the top of the search results (SERPS) without ever clicking through to your website. That’s not a temporary blip, it’s a structural shift in how search works. There’s every chance you are burning your Pay Per Click (PPC) budget.

Here’s another steer – don’t chase every new AI tool that lands in your inbox, an endless wave of shiny new products promising to revolutionise your digital presence. Instead go back to basics.

Review what you already have, identify the gaps, and fix your foundations first.

UK dealers are being targeted by an ever-growing number of AI marketing vendors, many of whom are selling solutions to problems dealers haven’t fully understood yet – I wrote an article called The Comprehension Gap. Understanding the landscape clearly is the essential first step. Some solutions don’t even need AI to resolve them.

The Language You Need to Know

To make sense of what’s happening, it helps to get comfortable with a handful of key terms that are increasingly shaping how dealers think about their digital presence.

SEO (Search Engine Optimisation) is the practice most dealers are already familiar with – building website content that ranks well in Google keyword searches and earns free, organic traffic. It remains important, but it’s no longer the whole picture. In fact Semrush, a leading analytics provider, recently  quoted a 1/3 drop off in search appearances for popular search terms because of ‘position zero’ AI summaries. Those enquiries are going to the forward thinking supplier who, ahead of the game in the world of GEO, is already being cited in AI search and AI App results – and will benefit from early adopter status forever.

Position Zero refers to the featured AI generated snippet that appears at the very top of Google results, providing a direct answer to a search query without the user needing to click any link. Appearing here means visibility without traffic, which is precisely the problem.

LLMs (Large Language Models) – as explained by Prof. Hannah Fry on the BBC this week – are the AI tools driven by state of the art electronic chips that are increasingly influencing how people search and make decisions. These include Google Gemini, ChatGPT, Claude, and Perplexity, platforms that millions of UK consumers now turn to for recommendations, comparisons, and buying guidance.

Schema Markup (sometimes called Website Schema) is structured data added to a website’s backend that consistently labels content, vehicles, reviews, services – so that search engines can properly understand it. It helps generate richer search results and improves click-through rates. Many dealer websites in the UK lack this entirely.

GEO (Generative Engine Optimisation) is perhaps the most important new concept for dealers to grasp. Where SEO is about ranking in traditional search, GEO is about structuring and creating content so it can be easily understood, cited, and summarised by AI tools. Think of it as SEO for the AI era. Unlike traditional SEO, which tends to return the same results for the same keyword, GEO is highly personalised, the same query from two different buyers can produce completely different AI-generated answers, pulling from reviews, forums, and third-party sources alongside your own website.

Prompt Visibility refers to how frequently and prominently your dealership, brand, or specific information appears in AI-generated responses. If an AI tool is recommending dealerships near a customer, does your name come up – and in what context?

Agentic AI is the next frontier. Unlike AI tools that simply provide information, Agentic AI can independently plan, make decisions, and execute multi-step tasks. In a retail automotive context, this means AI that could guide a buyer through the entire purchase journey, from initial research to booking a test drive, with minimal human involvement. This is not science fiction; early versions of this capability already exist. LLMs were initially created to drive Chatbot conversations.

Hallucinations are incorrect results appearing in AI search summaries that could affect a contact or even a loss of reputation. AI is in the ‘Model T Ford’ era, it is not always correct and you need to be testing search terms for accuracy as part of your reputation management.

What This Means for U.K. Dealers

The practical implications are significant. Every positive customer review matters more than ever, because AI tools draw on that content when forming recommendations. An authentic dealer with consistently strong, genuine reviews is far more likely to appear in AI-generated responses than one whose online reputation is thin or inconsistent.

Critically, responses to reviews also need to feel human and genuine – AI can detect overly automated or templated replies, and this affects how credibly your dealership is represented. You cannot rely on AI in most cases ChatGPT writing your content – it’s great for writer’s block, but rewrite ideas in your personal tone.

Your website content needs to answer real questions clearly and directly. Generic manufacturer copy doesn’t help AI understand what makes your dealership distinct. Pages that explain finance options, compare models, or provide genuinely useful local information are the kind of content that both search engines and AI tools favour.

From a technical perspective, your website needs to be accessible to AI crawlers in the first place. A significant number of dealer sites inadvertently block tools like ChatGPT and Google’s AI systems through robots.txt settings or security configurations – meaning those tools simply cannot recommend them, regardless of how good their stock or pricing is. Does your supplier understand llms.txt and is this included in your website back end? These are now questions you need answers to.

The fundamentals; fast loading websites, clean data, honest content, and a credible presence across all digital platforms haven’t changed. What’s changed is how much they matter, and what gets built on top of them.


Steve Coulter is a four decades Sales and Marketing expert with a career in the Automotive Industry and involved in state of the art Digital Marketing since 1999.

Technology: The Comprehension Gap

To stay abreast of technology and remain relevant in the sales and marketing industry I‘ve spent years mastering Web 3.0, the Metaverse, Cryptocurrency, and more recently Generative Engine Optimisation and AI Summary citations. The problem? These skills are useless if few people understand them. Businesses can’t adopt what they don’t grasp and in straightened times most won’t invest the time, money, or mental energy required to bridge that gap. The technology stalls and my consultancy stalls with it.

I’ve called it The Comprehension Gap. Given the title, I’ve attempted to be as concise as possible.


Every few years the internet reinvents itself.

Web3. The metaverse. Now AI. Each wave promises to change everything. Yet the same problem repeats: most people don’t really understand what it’s for.

That’s not a failing – it’s the point.

Mass adoption doesn’t happen because technology is clever. It happens when an average person can see the value without needing it explained. When the benefit is obvious, usage follows. When it’s not, adoption stalls.

Meta’s metaverse rebrand this decade is the sharpest example. Since 2020, it has burned more than seventy billion dollars building a future most couldn’t grasp. Not because the tech failed, but because the value wasn’t clear enough to justify the effort.

Most people didn’t reject it. They simply couldn’t answer a basic question: Why would I use this?

Web3 hit the same wall. Decentralisation, tokens, secure wallets, NFTs, blockchains. Technically clever, but hard to understand beyond enthusiasts. For most, the cost of understanding outweighed any benefit.

Here’s the uncomfortable truth:

  • If people need to learn new language before seeing value, adoption will be slow.
  • If using the tech feels like work, most will opt out.
  • If the benefit can’t be explained simply, it won’t spread.

This isn’t dumbing down. It’s how adoption works – remember Malcolm Gladwell’s description in The Tipping Point?.

The early internet succeeded because it didn’t ask people to change how they thought. Email felt like post. Search felt like asking a question. Social networks felt like a conversation. You didn’t need to understand the system. You just used it.

Web3 and the metaverse flipped this. They asked people to understand a fairly complicated premise first, then participate, then maybe benefit later. That order guarantees failure outside niches.

This gap between what technology can do and what average people grasp is expensive.

Big platforms pay for it. When users don’t “get it”, marketing becomes education, support becomes consultancy. Growth slows. Costs rise. Confidence evaporates.

Agencies pay too. Clients want to move forward but don’t understand what they’re signing up for. Strategy becomes explanation. Delivery becomes risk. Fees stop matching effort. Everyone feels busy; no one feels certain. An easy to confirm first page ranking becomes a ‘possible’ AI summary citation if the wind is in the right direction.

More truths:

  • When clients don’t understand the tech, they don’t trust outcomes.
  • When they don’t trust outcomes, they don’t commit long term.
  • When they don’t commit, adoption never scales.

Now AI and AI-driven search risk the same gap.

AI feels familiar. It fits on top of what we already do: write, search, research, plan. But the systems beneath are opaque. Rankings disappear. Attribution blurs. Decisions come from models that don’t explain themselves. The decision requires a leap of faith and businesses don’t run on faith.

For business leaders, that’s unsettling. Old rules vanish; new ones aren’t clear. Understanding lags behind capability.

When that happens, behaviour shifts. People hesitate. Budgets freeze. Vendors promising certainty gain power. Bad decisions follow.

This isn’t an innovation problem. It’s a comprehension problem.

Technology reshapes society only when it becomes boring enough to feel obvious. When people stop asking what it is and start asking how to use it. That moment won’t come if capability runs ahead of understanding.

The lesson from Web3 and the metaverse: the ideas weren’t wrong, but they weren’t simple enough for mass adoption.

AI won’t escape that fate unless it crosses the same line. The next internet phase won’t be won by the smartest systems. It will be won by the ones that make sense to non-experts.

Because mass adoption happens only when average comprehension is enough. Anything needing more will stay niche, no matter how much money pours in.

OPINION: Cars As iPhones

On a recent trip to the cinema I was presented with two new car adverts, both glossy. My cynical joke about the Range Rover that the owner was driving into the dealer for warranty work was soon overshadowed by the normalisation of a homogenised MG ‘SUV’ being £35000 as the basic price. But of course nobody pays the cash price nowadays, they mentally justify the purchase as multiples of their iPhone monthly payment. Cars as iPhones is a masterclass in behavioural psychology. 


Sometime in the last decade I described modern car buying as the iPhonication of the forecourt and if anything the idea has aged better than most of the vehicles now sold under it.

There was a time when the first question was the price. The full price. A number large enough to demand a pause and some moral arithmetic. Today nobody asks that. They ask how much a month, and the rest is treated as background noise, like terms and conditions or the weather.

The trick is simple. We no longer evaluate cars as capital purchases but as subscriptions. The monthly payment is quietly reconciled against the iPhone in your pocket. Fifty quid a month for a phone feels normal. Three hundred quid a month for a car becomes six phones. You already live with one. Why not six sitting outside?

This is not financial logic so much as consumer conditioning.

PCP and PCH did not just change how cars are paid for, they changed how they are justified. Ownership became vague, temporary, almost impolite to mention. What mattered was whether the number could slip into the standing orders without causing an argument. Once it could, permission was granted.

Electric cars pushed this model to its logical conclusion. On paper many are startlingly expensive. In practice they arrive softened by language about efficiency, tax, sustainability and the future. The monthly figure does the heavy lifting. Nobody emotionally processes fifty or sixty thousand pounds. They process four hundred a month by comparing it to phones, streaming services and a gym membership they forgot to cancel.

The iPhone model also smuggles in another assumption, that upgrading is natural and permanence is old fashioned. You do not keep a phone for a decade. You refresh it, guilt free, because you were never meant to own it outright. PCP borrows that psychology wholesale. Three years, hand it back, move on, feel modern.

Electric cars lean into this harder than most. Big screens, software updates, range improvements promised just around the corner. They are sold less as machines and more as devices, which makes their disposability feel progressive rather than wasteful.

The danger is that when everything is framed as affordable monthly, nothing feels expensive anymore. Stretch the term, adjust the mileage, tweak the deposit and almost anything can be made to behave. The decision shifts from judgement to tolerance. Not can I afford this, but can I live with it?

That is a profound change. It moves cost out of focus and replaces it with habit. Comparison does the rest.

‘The electric transition is often presented as a clean break from the past, but it has been powered by the most contemporary consumer mechanism of all. Permanent payment and planned obsolescence, wrapped in virtue and delivered by direct debit.’

The propulsion may be changing, but the mindset is familiar. We have not made cars cheaper. We have just learned how to stop looking at the whole number, and how to measure our lives in iPhones instead.