OPTIMISE

The AI Search Newsletter For UK Automotive – Issue 01

Welcome to the first issue. This newsletter has one job: to tell you what is happening in AI search as it relates to UK automotive, from independents and dealer groups to factory OEM, and to translate the US market a few months ahead of us so you can see what is coming before it arrives. Until the UK catches up the more interesting data will probably be from US sources.

OPTIMISE is also to inform and educate about the shift in Internet search that is already here and working against the traditional ways of being found online.

We are not covering AI chatbots in the showroom, voice agents in the call centre, or automated photo editing. This one is about a narrower and, we think, far more urgent question: when a buyer asks ChatGPT, Gemini or Perplexity where to find a car, will your brand or dealership be part of the answer and why.

1. Are You Visible? – This Month’s Spotlight

Every issue, we put real buyer questions into the major AI engines and report back on what gets recommended, what gets cited, and who is missing entirely. Run consistently over time, this becomes something most publications in this space do not have: a proprietary, longitudinal record of how AI search actually behaves towards UK automotive, month on month, rather than a one-off snapshot.

Actual intel you can employ. We will hopefully also see the shift when regular dealers have optimised for AI Search and not SEO for last year.

For this launch issue, we are setting out the format rather than running the full test. We wanted to get the test itself right first, because a single-town result only tells one dealer anything useful, and this newsletter is read by dealers nationwide.

The test splits into two parts, because they measure two different things.

Test A: what does AI recommend? This establishes what influences AI’s automotive answers in the first place, using product-level prompts like:

What’s the most popular 5 year old BMW to buy used in the UK?
What’s the most popular 5 year old SUV to buy used in the UK, any make?
What’s the most popular 5 year old EV to buy used in the UK?

Test B: where does AI say to buy it? This is the one that actually matters commercially, since the proposition of this newsletter is dealer and retailer visibility, not product trivia:

Where should I buy a five-year-old BMW in the UK?
Which BMW dealer is best for used BMWs in Sussex?
Who is the best independent used-car dealer near Littlehampton?

We will run both sets identically across ChatGPT, Gemini and Perplexity, and log more than just the answer given: the basis each response claims (sales data, reliability reputation, depreciation, or nothing stated), the type of source it appears to draw on (marketplace listing, manufacturer page, review site, forum, or none cited), whether it names a retailer or type of retailer, and whether the three engines agree with each other.

That gives every independent dealer, regardless of postcode, two separate and genuinely useful findings: which specific used stock is most likely to get recommended by default, and, more importantly, what it actually takes to be named as somewhere to buy from.

What we already know from the US market gives a strong hint at the pattern to expect. A mid-sized Ford dealer in mid-Michigan, working with a modest search budget and no increase in spend, sold 12% more vehicles in a flat market during 2025, largely on the back of a long-term SEO strategy built around Google reviews and third-party platforms rather than paid media – it turns out, exactly where AI Agents seek trust and entity signals. That is the pattern worth watching for in the UK: visibility built on reputation signals and consistent citations, not advertising spend.

Next issue: full results from both tests, logged against source type and reasoning given, plus a short list of quick fixes based on what we find. OK, you’ve read this far. We lied and have already run the prompt test with an interesting and useful outcome – to be published soon because why wait to be more visible?

2. The Number

Each issue, we lead with one figure worth knowing, and what it doesn’t tell you, since that second part is usually more useful than the number itself.

8.5%

That’s the share of visits to US dealer websites that Fullpath, an automotive customer data platform, attributes to AI-chat sources (ChatGPT, Gemini, Perplexity, Claude and Grok) as of April, having grown steadily through 2025 with a 43% jump in a single month in September and a further 34.5% surge in March 2026.

What it doesn’t tell us: it isn’t a UK figure, it isn’t industry-wide (this is Fullpath measuring its own customer base, not an independent census), and the detection method is largely referrer-based, which is known to undercount some AI traffic. Treat it as directional, not definitive.

Separately, and using a different method entirely, Ekho’s own buyer research found that 30% of shoppers used a generative AI tool at some point during their vehicle research journey. Two different companies, two different methods, pointing the same way.

The UK read: we do not yet have an equivalent published figure for British dealer websites. That gap is itself a story, and one we intend to start closing with our own tracking data in a future issue. But the use of generative AI is real and being used by your customers particularly in the research phase of a purchase.

3. AI Citation Sources & Web Page Structured Data

This is where the practical work happens, and where OEMs and dealer groups tend to diverge.

Manufacturers generally do well here because their spec sheets, model pages and press materials are already structured and consistent, which gives AI systems clear, machine-readable information to work with. Dealer groups and independents tend to struggle more, because visibility depends heavily on review platforms, local citation consistency, and third-party listing sites rather than the dealer’s own website.

A concrete UK example worth watching: Auto Trader’s AI-powered CoDriver tool has now passed 10,000 retailers using it, a sign that dealers are increasingly comfortable letting a third-party platform manage part of their digital presence. That is convenient in the short term, but it also means more of a dealer’s visibility sits on someone else’s platform rather than their own site, which is worth thinking through carefully rather than simply accepting.

Something to check this month: does your Google Business Profile, your main website, and your listings on Auto Trader and other marketplaces all describe your dealership consistently, using the same names, locations and specialisms? Small inconsistencies here are one of the most common reasons a business gets left out of an AI-generated answer.|

News: UK Watch – Click Dealer Puts Stock Inside ChatGPT (with a caveat)

While drafting this issue, we checked in on Click Dealer, part of automotive fintech group Centeca. What we found is worth flagging on its own terms: AI search advertising resembling Google’s vehicle listing ads, with inventory feeds appearing directly inside a chat response, is not a future possibility. It has already launched here in the UK.

In August, Click Dealer released a ChatGPT plugin connecting buyers with more than 120,000 vehicles from over 1,800 independent UK dealerships. A shopper can describe what they need in ordinary language, budget, family requirements, fuel type and intended use included, and the plugin surfaces live matching stock before handing the customer over to the supplying dealer’s own website.

But there is a caveat. This depends on a shopper already knowing the plugin exists and choosing to use it inside ChatGPT, or on ChatGPT itself surfacing it via its own connector recommendations, neither guaranteed. No independent figures have been published on how many buyers have actually used it since launch, so treat the reach of this as unproven for now rather than assumed.

It is a genuinely useful piece of engineering, and worth understanding. But it is worth being clear about what it is and is not.

This is feed-based placement. A dealer’s stock appears inside ChatGPT because they are connected to a structured product feed, in much the same way Google Vehicle Ads work. It gets a dealer into a specific transactional flow when a buyer is already searching for a car with defined criteria.

Important:

It is a different thing entirely from being properly cited by an AI model for the right reasons. A citation happens when the model has independently judged a business as credible and worth recommending, based on genuine signals such as consistent reviews, structured website data, and third-party validation. That is what determines whether a dealer appears in the much larger volume of open, conversational questions, such as “which dealer near me is best for family SUVs,” where there is no product feed to plug into at all.

A dealer group could join every feed-based scheme available and still be invisible the moment a buyer asks a genuinely open question. The two are separate jobs, and one does not substitute for the other. Feed placement is worth doing. It should not be mistaken for AI search visibility.

4. The Signal

Each issue, we introduce or apply one of the research concepts underpinning this newsletter’s methodology, and show what it means for a live automotive scenario.

This month: the Zero-Click Economy, from our Your Business Is Becoming Invisible report. A zero-click search is one where the user gets a satisfactory answer in the AI Summary results and never clicks through to any website. It is not a new idea, but AI has pushed it from being an occasional outcome to being in many cases of a simple answer the default one.

Across all Google searches, 68% now end without a click. When an AI Overview appears above the results, that rises to 83%. In Google’s AI Mode, it reaches 93%.

The distinction that matters most here is between ranking and being cited. A dealership can hold position one on Google and get nothing from it if an AI Overview answers the question above that result. A dealership whose content is structured to be cited, rather than merely ranked, appears inside the answer itself, which no traditional ranking position can replicate – think Page 5 position 6.

Car buyers spend an average of twelve hours researching before they ever visit a dealership. That research phase is exactly where AI systems are now doing the summarising, comparing and recommending, often without sending the buyer to any dealership website at all. A dealer who ranks well but is not structured to be cited is, for a large and growing share of that research phase, simply not part of the conversation.

UK automotive retail is still at an early stage of recognising this. Very few dealer groups have separated “how do we rank” from “how do we get cited” as two distinct problems requiring two distinct approaches, and that gap is exactly where the coming issues of this newsletter will keep returning.

5. 12 Months Ahead

The US runs ahead of the UK on AI search behaviour, typically by 12 to 18 months. Each issue, one story from across the Atlantic, filtered strictly to AI search rather than AI operations, and what it means for the timeline here.

What happened: Ekho, a US company, has launched a dedicated benchmarking product that lets dealers compare how they appear in AI search results against named competitors.

Why it matters: AI visibility tracking is moving from a niche consultancy service into a proper SAAS (Software as a Service) category. That shift changes who a dealer group turns to first when they want to know where they stand. An expert may offer more accurate advice than a website provider.

When we expect it here: watch this space. If the US pattern holds, a UK equivalent becoming commercially available is plausible within the next year. We are very well connected in the digital search SAAS space so may be able to white label a forthcoming product.

What happened: US commentators had been speculating that AI search advertising could come to resemble Google’s vehicle listing ads, with inventory feeds appearing directly inside a chat response.

Why it matters: as it turns out, the UK got there first. See the Click Dealer story above for what this actually looks like once it exists, and why it is not the same thing as AI search visibility.

When we expect it here: already live. This is the rare case where the UK is ahead, not behind.

One to watch each month: Fullpath, Ekho and Cox Automotive all publish regular US data on this topic. We will keep drawing from these three consistently so you get a reliable trend line rather than one-off stories.

That’s Issue 01. If there’s a UK dealer group, OEM campaign, or piece of AI search data you would like us to look into, message to say hi and we will follow it up for the next issue.

About OPTIMUM

OPTIMUM is State of the Art Digital’s AI SEO and GEO audit framework, scoring a business across seven dimensions of AI search readiness with a Now Score, a Future Score and the Delta between them. It is the methodology behind the research and terminology referenced throughout this newsletter.

About AI Search, Today and Tomorrow

AI Search, Today and Tomorrow is State of the Art Digital’s combined research study, bringing together three flagship reports on how AI search is reshaping business visibility, including the Zero-Click Economy findings referenced above. It is updated and republished on a quarterly basis as the AI search landscape develops. A copy is available on request.

OPTIMISE is written and researched by Steve Coulter, founder of State of the Art Digital. To get in touch e-mail steve@stevecoulter.co.uk