How Is AI Changing the Way Customers Choose a Car? The Automotive Customer Journey in 2026

Buying a car still largely ends offline, during a test drive, a conversation with a sales advisor or when the financing is finalized. What’s changing, however, is what happens before that. Before visiting a showroom, a customer may do multi-step research on Google, watch road tests on YouTube, read reviews and ask ChatGPT, Gemini or another AI tool about the best model for their needs.

The scale of this change keeps growing. In a 2026 Invoca study, 63% of automotive consumers said they had used generative AI while researching a car purchase. A year earlier, it was 41%.

That doesn’t mean AI is replacing people, though. 86% of respondents still consider human contact important when buying a car.

This changes how we look at the automotive customer journey. A brand has to be present not only when the user already knows its name. It also has to show up when the customer is only just wondering what kind of car they actually need.

Just a few years ago, the customer journey could look relatively simple:

Today, users rely on many sources at once:

The biggest change isn’t necessarily happening at the very end of the journey, but during discovery and consideration, when the customer is only just defining their needs and choosing the brands and models they want to look into further.

This is an important change from an SEO perspective too. A brand can be added to or removed from the shortlist before the user even types its name into a search engine.

Users don’t have to start with the question: “Toyota RAV4 or Kia Sportage?” They can describe their problem first:

• Which family SUV under $50,000?
• Which hybrid car works best in the city?
• Which car should I buy if I drive about 19,000 miles a year?
• Which EV is good for long trips?
• Which car will work well for a family of four?
• Which car has a large trunk and low fuel consumption?

In this situation, the user doesn’t name any brand. It’s AI that may suggest specific models and manufacturers.

This is one of the reasons why brand visibility means something different now. In classic SEO, a brand mainly competed for a ranking and a click. In the AI environment, it also competes to make it into the answer and onto the shortlist at all.

Imagine a user who types into ChatGPT:

“I have $43,000, two kids, and I drive about 12,000 miles a year, mostly on the highway. I’m looking for an SUV with an automatic transmission, a large trunk and the lowest possible fuel consumption. What should I choose?”

This is no longer a classic single-keyword search. The user gives AI their budget, family size, annual mileage, how they’ll use the car, the body type, the preferred transmission and their fuel consumption expectations.

Based on this, the model can create a shortlist of several cars and explain why they’re worth considering. This consideration set is becoming one of the most important points of the new customer journey.

In addition, a BrightEdge analysis shows that in ChatGPT, around 70% of transactional automotive queries were still non-branded. This means that even users with very high purchase intent may not start with a specific brand.

This doesn’t mean the end of classic SEO. It does mean, however, that the road to a branded search may start earlier and run through other sources.

Before generative AI arrived, a user might run a series of micro-searches: Toyota RAV4 fuel consumption, Kia Sportage fuel consumption, Toyota RAV4 trunk, Kia Sportage trunk, RAV4 vs Sportage, RAV4 reviews, Sportage reviews. Each query led to another stage of research.

Today, part of this process can fit into a single prompt:

“Compare the Toyota RAV4 and the Kia Sportage in terms of fuel consumption, trunk space, running costs and suitability for a family of four. Tell me which model would be the better choice.”

So AI can compress many micro-searches into one conversation. That doesn’t mean users stop verifying information, though. Quite the opposite: after getting a shortlist, they may move on to Google, YouTube, manufacturers’ websites, reviews or dealer offers.

AI helps narrow down the choice → Google and the media help verify it → the manufacturer provides detailed information → the dealer helps turn it into an actual purchase.

The automotive customer journey isn’t moving entirely to ChatGPT or Gemini. Users may turn to different sources depending on what they need at a given moment.

This matters because the digital automotive customer journey was already very complex. In one of Google’s older studies of this market, users made more than 900 digital interactions before reaching a decision.

So AI isn’t creating the customer journey from scratch. What it can do is shorten the path between a question and a specific recommendation.

Yes. And this is exactly where it’s worth being careful not to draw far-reaching conclusions about automating car purchases.

A 2025 CarGurus study showed that 83% of customers would like to complete more steps of buying a car from home. At the same time, 86% still want to see the car in person before making a decision.

This shows well that digitizing the process doesn’t mean moving it entirely online. AI can help a customer decide: “These three models best fit my needs.” The dealer, in turn, can answer the question: “Which one can I see tomorrow, what’s its price and can I take a test drive?”

That’s why it’s worth seeing both areas as two consecutive stages of one journey rather than as competing strategies. We write more about this in our article Car Dealer vs. Manufacturer. Which Marketing Activities Deliver Results?

This is one of the more practical problems with measuring AI’s impact. Imagine the following path:

ChatGPT recommends a model → the user searches for the model on Google → watches a road test on YouTube → visits the manufacturer’s website from organic search → checks a local dealer’s offer → calls → books a test drive → buys the car

In Google Analytics 4, we won’t necessarily see ChatGPT as the source of this conversion. That’s why two things need to be kept apart: AI as a traffic source and AI as a touchpoint that influences the decision.

Referral traffic from ChatGPT can be easy to measure. It’s much harder to measure a situation where AI influenced the choice of model, but the user later came to the brand through Google.

That’s exactly why it’s worth monitoring not just traffic but also brand visibility, share of results (Share of Voice), citations, mentions and presence in answers to the most important purchase prompts.

Proprietary data is particularly important because it shows what these changes look like in a specific market.

Delante’s study analyzed 10 automotive brands in terms of organic visibility, the number of referring domains, AI citations, mentions, Share of Voice, answers to high purchase intent prompts and car registrations, among other things.

Of course, this doesn’t mean more links automatically lead to more sales or more citations. Correlation isn’t proof of causation. The result does show something important, though: a brand’s digital authority and its presence across different search ecosystems don’t function as completely independent areas.

In the analyzed set, the five brands with the most AI citations, Toyota, BMW, Audi, Mercedes and Skoda, accounted for more than 40% of registrations.

Automotive isn’t just something we analyze and write reports about. It’s also an industry where we’ve actually run SEO and SEM campaigns.

If a customer can ask AI about a car they don’t know yet, having extensive model pages isn’t enough on its own. Manufacturers should also answer the questions that come up before a brand is chosen.

This isn’t about creating hundreds of generic articles, though. The content should provide specific, verifiable information that answers users’ real needs. That also increases the chance that the brand will appear in external publications, comparisons and roundups. We write more about building this kind of visibility in our article How to Build Links That Improve Your Visibility in AI Models?

For dealers, Local SEO and the availability of information become especially important. A customer who has already received a model recommendation from AI may take the next step: “Where can I see this car near me?” or “Which dealer has this model available right now?”

At this stage, completely different information matters than during discovery. The customer wants to know where the showroom is, whether a specific model is available, what the current price is, whether they can take a test drive, what the financing options are, what customer reviews say, how to contact the showroom and when it’s open.

That’s why the Google Business Profile, local landing pages, the current offer, reviews and contact details aren’t an add-on to the strategy. They’re part of the final stages of the customer journey.

This is where it’s easy to fall into a narrative trap: SEO is ending → now it’s time for AISO. That’s too much of a simplification.

SEO is still responsible for the technical and content foundations of visibility, including architecture, indexing, crawlability, content and authority. GEO covers activities related to visibility in the Google environment, including the search engine’s generative features. AISO extends the perspective to environments such as ChatGPT, Gemini and Perplexity.

These aren’t three independent worlds, though.

Nie trzeba zaczynać od przebudowania całej strategii. Warto najpierw sprawdzić, jak marka wygląda w konkretnych scenariuszach zakupowych.

This is the moment that may decide which models later end up on their shortlist.

Automotive customers still need a car, a test drive, financing and a person on the other side. What has changed above all is what happens before that.

A manufacturer no longer competes only for rankings on a phrase related to a specific model. It also competes for its car to show up when the customer asks AI: “What should I buy?”

Dealers, in turn, don’t compete only for the phrase “car dealership + city”. They also have to be easy to find when the user wants to move from a recommendation to a specific car, its availability and a test drive.