Your point that nobody has agreed on how to measure this is the part I would build on. Split the visibility number into three series instead of one score: does the answer name the brand, does it link a page, does it list the brand as an option. They move independently and one score hides which changed. Pair them with an outcome the business already owns, and freeze the prompt set for a quarter so engine churn is not read as your own progress. Then trace one prompt backwards to the third-party page that changed, and you have your earning mechanism.
P.S. still such a smart move from Rand to start Sparktorro, so much ahead of its time!
Your point that nobody has agreed on how to measure this is the part I would build on. Split the visibility number into three series instead of one score: does the answer name the brand, does it link a page, does it list the brand as an option. They move independently and one score hides which changed. Pair them with an outcome the business already owns, and freeze the prompt set for a quarter so engine churn is not read as your own progress. Then trace one prompt backwards to the third-party page that changed, and you have your earning mechanism.
P.S. still such a smart move from Rand to start Sparktorro, so much ahead of its time!
can you give an example of original data/proprietary framework that AI loves citing the most
Anything talking about trends and pain points is usually picked up very well by LLMs if you market the studies well. I won't disclose our clients' assets (for others not to copy the strategy) but we have some cool case studies here: https://www.smarty.marketing/wp-content/uploads/2026/04/Digital-PR-service-case-study.pdf