Helping JLR understand why Range Rover wins or loses in AI and where to act

Buying a Range Rover involves substantial research. Customers compare vehicles, weigh up specifications, needs and costs before making a significant financial commitment.

JLR recognised that AI assistants were becoming part of that process. A prospective customer could ask ChatGPT which luxury SUV was best for a family of 5 that can easily tow a horse box, compare Range Rover Sport with a BMW X5 or seek advice about the differences between vehicles in the Range Rover line-up.

The business wanted to understand how its vehicles appeared in those conversations and what it could do to improve their representation.

JLR already had established technology and agency relationships, including Adobe and Accenture. The gap, as JLR described it to Zebora, was in the diagnosis and strategic recommendations. The marketing leadership team wasn't being told clearly enough on what they should do. So they needed help understanding why particular vehicles were being recommended (or not), where opportunities were being missed or which actions should take priority and why.

JLR also wanted a way to explain the AI opportunity to its board in a way that could easily be understood and connected AI recommendations to familiar marketing decisions about product positioning, content and PR.

The initial engagement focused on Range Rover in the UK, using ChatGPT to establish a baseline and investigate what was driving its performance.

A straightforward brand mention count would have missed much of what JLR needed to know.

Range Rover has several vehicle ranges with different propositions and competitors. It also has a flagship range 'Range Rover' which shares the same name as the brand.

We noticed that AI responses do not always describe car brands consistently. ChatGPT might recommend Range Rover as a brand, name Sport or Evoque, identify a specific powertrain or refer to a vehicle that is no longer sold. It could also recommend a model available in the US but not in the UK, even when the query was run from a UK account.

The same variations appeared in responses about competing manufacturers.

JLR needed to see how Range Rover performed overall against BMW, Mercedes and Porsche while understanding how its individual ranges of vehicles compared with relevant alternatives. It also needed to know which customer needs were prompting recommendations and where Range Rover was being overlooked.

That required a measurement approach built around the vehicle portfolio rather than a generic set of automotive search terms.

Measurement built around the business

Zebora worked through Range Rover's portfolio, customer needs and competitive landscape to develop a bespoke framework of 759 prompts across 80 reporting nodes.

The prompts reflected different stages of vehicle research, from broad discovery to comparisons and questions about specific models. Multiple prompts were aggregated within each reporting node, with commercial weighting used to give greater importance to relevant customer needs and opportunities.

Zebora also adapted its platform to map inconsistent vehicle names and variants into a meaningful automotive hierarchy. JLR could then examine performance at brand level or look more closely at individual ranges, including how Range Rover Sport and Evoque compared with their competitors.

Investigating what was behind the results

The ChatGPT analysis showed where Range Rover appeared, how it was positioned and which customer needs it was missing.

Zebora then audited the Range Rover homepage and three vehicle range pages to examine how clearly the website explained each model, who it was for and what made it distinctive.

Looking at the AI responses alongside the website content helped the team identify plausible reasons for some of the gaps and turn the diagnosis into specific recommendations. The website findings were a useful part of that explanation, although they did not establish that website content alone caused the AI responses.

Strong brand recognition was masking weaknesses across the vehicle range

Range Rover's initial ChatGPT visibility and sentiment both scored 77/100. It ranked third behind BMW and Mercedes in the competitive analysis, with its flagship model contributing disproportionately to the wider range's performance.

ChatGPT understood Range Rover's heritage, luxury credentials and off-road capabilities. It was less clear about the individual propositions of Sport, Velar and Evoque and how these were differentiated from the better understood flagship range.

That lack of distinction showed up in recommendations for particular customer needs. Range Rover was missing opportunities in areas including family use, urban driving, daily commutes and safety.

For JLR, the implication was practical: stronger model-specific messaging could help prospective customers, and potentially AI systems, understand which vehicle suited which need.

The findings were presented to JLR's board by its Global Group Marketing Director. The response included a question about how the approach could be replicated across Jaguar and Defender.

JLR subsequently commissioned an expansion to replicate what we'd done for Range Rover across the Jaguar and Defender brands in the UK. We were also to include additional AI platforms, including Google AI Overviews, Claude and Perplexity.

The teams also explored how Zebora could support future car/SUV launches by preparing a clear LLM roadmap that coordinates product messaging and relevant third-party signals before a new model reached the market. JLR expressed interest in Zebora's PR module, which connects important visibility gaps with the prompts and sources behind them so that PR teams can focus their efforts.