What the Co-Mentions page shows
Co-Mentions shows you which other companies appear in the same AI responses as your brand. It surfaces patterns across thousands of AI-generated answers, revealing who AI models group you with. Whether that's direct competitors, complementary tools, or unexpected peers.
Understand how co-mentions impact your AI positioning
When an AI model answers a question about your category, it doesn't name one company, it names several. The brands that consistently appear alongside yours are a strong signal how AI perceives your positioning. If you're always co-mentioned with legacy incumbents, you're being framed as a legacy player. If you appear alongside fast-growing alternatives, you're being framed as an emerging option. Understanding your co-mention landscape is the first step to actively influencing it.
How ALLMO calculates brand co-occurrence using Jaccard Score and Lift
ALLMO analyzes every AI response collected for your report and identifies all companies mentioned. For each brand that co-occurs with yours, two statistical metrics are calculated:
Jaccard Score measures the overlap between response sets: how often both your brand and theirs appear in the same responses, relative to how often either appears at all. A high Jaccard score means tight association.
Lift measures whether you appear together more than chance would predict. A lift above 1 means the co-occurrence is meaningful. The higher the lift, the stronger the signal that AI models treat you as related.
As always, you can filter by date, AI model, language, country, and tags. Further, you can set a minimum mention threshold to remove low-frequency noise from the analysis.
How to sort, filter, and search the co-mentions table
Open Co-Mentions from the Explore section
Search for a specific competitor to check whether and how often you're grouped with them.
Use the Min. mentions slider to set a meaningful floor. Start at 5 or 10 to filter out one-off appearances.
Review the table sorted by Lift (descending) to find the strongest non-random associations.
Use the Jaccard Score to identify brands that are deeply associated with yours across the widest range of responses.
What to do with the insights
If you're frequently co-mentioned with a direct competitor: Your brand and theirs are being treated as very similar, or at worst interchangeable, alternatives. Review what differentiates you and ensure that differentiation is explicit in your web content, product pages, and any sources AI models are likely to cite.
If your lift scores are low across the board: Your brand may be appearing in isolation, which can indicate strong standalone visibility, but may also mean that AI models don't yet have a clear category frame for you. Creating content that explicitly compares you to known alternatives can help establish context.
If unexpected companies appear with high lift: Investigate why. AI models may be drawing from sources that frame your brand in a context you aren't aware of. This is useful competitive intelligence, and a signal worth acting on.