English 
Todonada Co., Ltd., which operates the PR effectiveness measurement tool “Qlipper,” has added three functions to its LLM (Large-Scale Language Model) brand awareness measurement tool “DigiDigi” : Automatic creation of monitoring projects. (Qlipper integration)
Background: The next question after “Is it recalled?” is “Is the recall set robust or unstable?”
With the proliferation of generative AI, consumers are increasingly using LLMs (Limited Liability Managers) as a starting point for searching for products and services. The new standard for PR and public relations is whether a brand is included in the “recall set” —the set of brands that appear in the LLM’s responses when asked, “What beer do you recommend?”
Our intergenerational comparative study of the LLM model, published in August 2026, confirmed that recall sets change within short generational cycles, and that the stability of recall sets varies significantly depending on the category. In categories with strong recall sets, reversing the hierarchy is not easy, while categories with unstable hierarchies offer room for entry. Understanding which state your company’s category is in is the starting point for public relations strategies in the LLM era.
こちらもお読みください： 「Fusion Connect」は、企業におけるAIの導入を容易にします
This new feature allows anyone to check the “strength of the hierarchy” with a single measurement, and furthermore, it connects the measurement results directly to the next action (continuous monitoring, media approach).
出典： PRtimes
日本語