Useful split between position and presence. One pitfall when you add the prompt-set layer: AI answers vary run to run, so a single check can mislead. Run each prompt 3 times and report the mention rate (e.g. 2 of 3) rather than yes/no, and keep the prompt wording frozen month to month so the trend is comparable. In GA4, a custom channel group matching referrers like chatgpt.com, perplexity.ai and gemini.google.com keeps AI visits separate from generic referral. For the client report, one extra column that helps: which competitor was named when you weren't. It tells the content team exactly which comparison page to write next.
Thanks Charlie, these are great additions. The run-to-run variance point is the one people skip most. A single yes/no check makes a dashboard look more precise than it is, so 3 runs and a mention rate is a much more honest number, and freezing the prompt wording is what makes the month-to-month trend actually mean something. The "which competitor was named instead" column is a smart one too. It turns a visibility report into a content brief, which is usually what the client really wants from it.
Charlie from xSquads
Head of GTM @ xSquads (xsquads.ai)
Useful split between position and presence. One pitfall when you add the prompt-set layer: AI answers vary run to run, so a single check can mislead. Run each prompt 3 times and report the mention rate (e.g. 2 of 3) rather than yes/no, and keep the prompt wording frozen month to month so the trend is comparable. In GA4, a custom channel group matching referrers like chatgpt.com, perplexity.ai and gemini.google.com keeps AI visits separate from generic referral. For the client report, one extra column that helps: which competitor was named when you weren't. It tells the content team exactly which comparison page to write next.