Plain answers
Blog
Each post answers one question a real owner asked, in the first two sentences, then explains from work we did. Sources are named at the bottom of every post.
The blog answers the questions owners ask before they pay anyone for search or AI visibility work: whether a file or a tactic they were quoted for does anything, what happens to a site when they leave a platform, who owns the domain and the profile, how to check which businesses an AI names in their town, and whether anyone can promise a spot in an AI answer.
Two rules apply to every post. First, every post answers one question a real owner asked us, and the answer is in the first two sentences, before any explanation. Second, every statistic is sourced and linked. When a post says Google does not use a file, it links to the page where Google says so. When it cites a survey, it names the survey, the sample and the date. A claim we cannot link to a source does not go in.
The posts are written from work we did, not from other people's summaries of it. Where the answer is no, the post says no, and says what to spend the money on instead. New posts are added when there is a question that needs a page, not on a schedule.
Does my business need an llms.txt file?
No. Google has published that its Search and AI features do not use llms.txt. What the file is, why agencies sell it, and what to spend the money on instead.
What an AI wrote when it picked our firm
A client sent us the memo an AI assistant wrote when it chose Insight Investigative Group from an 11-firm shortlist. Seven checks, each one a fact on a page.