SEO and AISO for Car Models: How to Get Google and AI to Recommend Your Car

You need to distinguish between two groups of queries that a model page answers today.

Brand queries capture demand that already exists: the user knows what they’re looking for and searches for a specific brand. With non-brand queries, SEO and AISO help you get into the consideration set, the set of models taken into account before a decision is made.

A spec sheet should answer the question “what does this car have?” Good content optimization, on the other hand, should answer the question “who does it matter to?”

A table that just says “trunk: 580 l” is less useful than data combined with an interpretation: “The 580-liter trunk easily fits a stroller and the whole family’s luggage, so the model works well for longer family trips.” Data describes the car. Context tells AI who to recommend it to.

Many automotive model pages look visually appealing, but from a search point of view they contain relatively little information: a hero section, a few photos, a configurator and a spec sheet hidden in a PDF. Unfortunately, these elements are poor sources of knowledge for search engines and AI.

That’s why the important information on the page should be:

• Crawlable: open to Google’s crawlers, meaning not blocked in the robots.txt file.
• Indexable: free of tags such as “noindex”, so it can actually get into search results.
• Text-based: written in the code as real text, not locked in an attached PDF price list or burned into photos of the vehicle.
• Available in HTML: visible to the search engine right away, without having to load and click through complex JavaScript-based configurators.
• Logically structured: organized with a clear heading hierarchy (H1, H2, H3).
• Connected through internal links: linked to other parts of the site to guide crawlers and users smoothly to, for example, comparison content, models available right away or straight to dealer pages.

Google officially stresses that the same fundamentals also apply to AI Overviews and AI Mode.

Schema.org has a dedicated Vehicle type that lets you describe a vehicle’s specifications. Google also recommends using Product data, and for cars it notes that, where appropriate, the Product + Car types can be used together.

Google has a dedicated data standard for car listings, called Vehicle Listings, which uses, among other things:

• make
• model
• price
• VIN
• condition
• mileage
• availability
• MSRP
• a link to the specific vehicle

Keep in mind that not all Vehicle Listings features are available in every country (some aren’t available in Poland, for example). Still, it shows that search relies more and more on well-structured, unambiguous product data, so it’s worth preparing for this in advance.

In automotive, price lists and technical data usually end up in a PDF. The PDF can stay, but the most important data should live directly on the page as text.

Instead of just writing “Download the technical data,” it’s worth showing the information directly, e.g.:

• trunk: 650 l
• seats: 7
• towing capacity: X lb
• AWD
• range: X miles

We’ve already mentioned enriching data with context. In practice, the easiest way to turn that context into something useful is short, specific answers to the real questions customers bring to search engines and LLMs.

A model page can include sections that answer questions such as:

• Is this model a good fit for a family?
• How does it handle long trips?
• HEV or PHEV: which version should I choose?
• How much does the trunk really hold?
• Which version is right for a business?
• Is it suitable for towing a trailer?
• What’s charging like on the road?

This kind of content is closer to the way users phrase their questions to AI than a classic technical spec table. It doesn’t mean giving up the spec sheet, just adding to it.

Automotive is a comparison category by nature.

Users ask:

• X or Y?
• Which one has a bigger trunk?
• Which one uses less fuel?
• Which one is better for a family?
• PHEV or HEV?
• SUV or station wagon?

Manufacturers shouldn’t leave the whole comparison narrative to portals, the media or competitors. It’s worth creating your own comparison content at three levels:

1. Within the model: version A vs. B, e.g. comparing engine options or trim packages for the same model.
2. Within the brand: model X vs. model Y, helping customers choose between different models from the same manufacturer.
3. Within the category: SUV vs. station wagon, EV vs. HEV, comparisons at the level of the whole segment, regardless of brand.

The manufacturer is, of course, the best source of information on specifications, prices and equipment, but for a question like “which SUV is the best?” users and AI also need external sources: tests, the media, rankings and expert publications.

AI may turn to the manufacturer’s website, but it may also pull data from an industry portal, a Reddit thread or a YouTube video. So the manufacturer doesn’t control the whole AI answer. It’s one of many sources the model builds its answer from.

That’s why the strategy shouldn’t stop at the question “how do we optimize the model page?” A better question is: how do we build a whole ecosystem of information about the model that AI can rely on?

That ecosystem is the car’s digital footprint, the trail a model leaves online beyond the manufacturer’s own website.

It’s made up of, among other things:

• dealers
• the media
• tests
• experts
• video content
• rankings
• mentions
• links pointing to all of these sources

You can read more about how AI is changing the way customers choose a car on our blog.

The model page on the manufacturer’s website and the page for a specific vehicle at a dealer, the so-called Vehicle Detail Page (VDP), answer completely different customer questions.

How Does Content Differ on the Manufacturer’s and the Dealer’s Websites?

The manufacturer should answer the question: is this car right for me? The dealer: where can I buy it?

This distinction translates directly into strategy. We wrote more about how the two segments of the market approach this in our article Car Dealer vs. Manufacturer.

Ahrefs has shown with its own data that AI systems can mention a brand many times without linking to it. That’s an important reporting trap: if you measure a model’s visibility in AI only by the number of visits coming from ChatGPT, you lose sight of most of the situations in which the model was actually mentioned and influenced the user’s decision.

So a model’s visibility in AI doesn’t come down to the question “does ChatGPT cite the BMW X3?” A better question is: does the BMW X3 show up for queries about a premium family SUV, an SUV for the city, a car for long trips and an alternative to model Y?

Which Key Performance Indicators (KPIs) Should You Monitor?

In our report on automotive brands’ visibility in AI, we set data on organic traffic, link profiles and brands’ presence in LLMs against car registration data. One of the strongest results was a 0.89 correlation between the number of referring domains and brand citations in AI.

This doesn’t mean a single link will automatically make ChatGPT recommend a car more often. It does show, however, that a model’s presence beyond its own website, and digital authority in the broad sense, are something an automotive AISO strategy can’t ignore.

Delante also works in the automotive industry, and we’ve described the results of our partnerships in more detail in our automotive case study.

By the time a customer types in the name of a specific model, most of the decisions about which cars they want to consider may already have been made. That’s why, in automotive, it’s no longer enough to optimize your branded phrases well.

A car model should be understandable to search as an answer to a specific need: family, budget, driving style, range, space, safety or financing options.

SEO makes information accessible, organized and findable, while AISO adds a further question to that process: does your car also show up when the user isn’t asking about your brand yet?

Because in 2026, the winner isn’t only the model that ranks first for its own name. It’s also the one that makes the shortlist before the customer has decided what they’re actually looking for.