Google AI Overview Places Utah Companies Earlier Than Other Engines
Average mention position ranges from 1.92 on Google AI Overview to 2.71 on ChatGPT, meaning recommendation quality varies beyond simple visibility.
2.44 average position
The finding
Across 382 total appearances, Utah AI companies appeared at an average position of 2.44 in recommendation lists. But this average obscures meaningful variation between engines.
| Engine | Average Mention Position |
|---|---|
| Google AI Overview | 1.92 |
| Gemini | 2.48 |
| Claude | 2.6 |
| ChatGPT | 2.71 |
Google AI Overview places Utah companies earliest, nearly a full position ahead of ChatGPT.
Why position matters
AI recommendations are typically ordered. A model asked "best CMMS software" might return a list of five or seven vendors. Position one is seen by every reader. Position five may be skimmed or ignored entirely.
Traditional search taught us that click-through rates decay rapidly: position one gets dramatically more clicks than position three, which gets dramatically more than position seven. While we do not have click data for AI recommendations, the same attention dynamics likely apply. A recommendation is not binary. Being first is better than being fifth.
What explains the engine differences
Several factors may contribute to position variation:
Retrieval ranking: Each engine ranks retrieved sources differently before synthesising. If an engine retrieves a source that prominently features a Utah company, that company may rank higher in the final output.
Summarisation logic: Engines have different approaches to ordering recommendations. Some may default to alphabetical, some to source prominence, some to their own relevance scoring.
Category competition: The number and strength of national competitors in each category affects where regional players rank. Categories with fewer dominant national players may give Utah companies higher positions.
The five most visible companies
The 5 companies with the highest visibility across all engines and prompts were:
- Limble CMMS
- Recursion
- Spiff
- Weave
- Entrata
These companies not only appear frequently but tend to appear earlier in recommendation lists. They have solved both the visibility and the position problem.
What this means for companies
A company appearing at position four has a weaker foothold than one appearing at position one. The optimisation goal is not just "appear somewhere" but "appear early." This likely requires not just presence on third-party sources but prominent presence: being featured, reviewed positively, mentioned first.
The data also suggests that different engines may be more receptive to Utah companies. If your company consistently ranks higher on Google AI Overview than ChatGPT, that tells you something about where your existing authority signals are strongest. It may also inform where to focus content distribution, since Google AI Overview draws heavily from YouTube and Reddit while ChatGPT leans on Gartner and LinkedIn.
Other findings
Which Utah AI companies never appear in AI-generated recommendations?
Twelve Utah AI Leaders Are Completely Invisible to AI Engines
One in four companies from the AI Utah 100 list never appeared in any AI answer to any buyer-intent prompt across all four engines tested.
Does adding 'in Utah' to a search prompt help local companies get recommended?
Utah-Qualified Prompts Lift AI Visibility by More Than Half
When prompts specified Utah, appearance rates rose from 29.6 percent to 44.7 percent, suggesting geographic signals are a real lever for regional companies.
Do different AI engines recommend the same companies?
No Single AI Visibility Score Captures the Full Picture
13 companies (27.7 percent) appear on some engines but not others, meaning visibility depends on which AI assistant a buyer uses.
Which companies get recommended instead of invisible Utah AI companies?
When Utah Companies Are Invisible, These Competitors Win
For each of the 12 invisible Utah companies, a consistent set of national competitors captures the recommendations they do not.
Want to know how AI answers describe you?
We run the same measurement on your category. Fifteen minutes with founder Omar Jenblat, your own numbers, no deck.
