GEO: how to measure and manage your visibility across AI search engines?
- [ IA ]
- [ Meteoria ]
- [ GEO ]
The challengesAnticipating the rise of AI search engines in the customer journey
As the publisher of an IoT hypervision platform that centralises data from multiple areas of activity (energy optimisation, parking, asset management, street lighting, etc.), Kuzzle noticed that traffic from AI search engines was becoming increasingly significant as the number of ChatGPT and Perplexity users grew.
To avoid losing visibility, the company turned to Kaliop to measure its visibility across AI search engines and develop a roadmap to strengthen it.
“We knew that generative search was going to become an important topic for us, but we didn’t yet have the means to objectively measure our visibility across these new channels. Kaliop’s methodology enabled us to define a clear roadmap.
Successful KPIs
0
AI queries analysed, across around 40 prompts
0
overall brand visibility for Kuzzle across the queries analysed
0
brand sentiment score
Anticipating the rise of AI search engines in the customer journey
On its website, Kuzzle noticed that traffic from AI search engines was becoming increasingly significant as the number of ChatGPT and Perplexity users grew (the audit was conducted before Google AI Overviews was launched). Keen to maintain its visibility on this new channel, the company asked itself: does the Kuzzle brand appear when users ask AI about the topics it wants to be associated with? While most of its customers do not come directly from the internet, this digital visibility supports its sales efforts in the field and represents a growing share of acquisition. Kuzzle therefore asked Kaliop to answer three questions: how visible is the brand across AI prompts? What sentiment is associated with it (positive or negative)? And do competitors appear more prominently in the answers?
Measuring visibility across 2,500 AI queries
Kaliop and Kuzzle jointly defined the scope of the audit: 80 prompts targeting France, monitored across ChatGPT and Perplexity. Using the Météoria tool, each prompt was simulated between 30 and 40 times to build a statistically representative dataset: more than 5,000 queries were analysed in total, smoothing out variations in responses from one AI model to another.
For each business topic (energy optimisation, parking, asset management, sanitation, the environment, climate, public lighting, energy, etc.), the tool calculates three indicators:
- Visibility rate: the percentage of queries in which the brand is mentioned;
- Reach: the number of prompts that generated a mention of the brand;
- Average brand position: the brand's average position among the brands mentioned in the responses.
Three levels of visibility are distinguished: high (1st to 3rd position), medium (4th to 8th/9th position), and none (brand absent from the responses).
Another key aspect of the audit was identifying the sources used by AI search engines, in order to develop a GEO strategy across external websites. This information turns the audit into an actionable plan: to increase visibility on an AI search engine, a brand needs to be present on the websites it uses as sources.
Turning AI sources into an editorial action plan
Based on the sources identified, Kaliop developed a prioritised content roadmap for Kuzzle, organised by topic. Six areas in which the brand needed to strengthen its content and visibility were identified as priorities.