AdTech company MarkApp says it launched Shelf on 14 November, positioning the product as a way to measure which brands, products and retailers AI assistants recommend in response to shopping questions. The ExchangeWire PressBox item appeared on 13 November, while the announcement itself names 14 November as the launch date. The service targets brand visibility, reputation and media-planning teams.
According to MarkApp's release, the service uses a fixed monthly panel of 50 to 100 real customer questions. Each question is sent to assistants through their official APIs three times with web search enabled. Shelf records which brands are named, their order in the answer, the supporting reasons and the cited websites.
The company advertises an AI Visibility Score from 0 to 100, share of answers compared with competitors, incorrect product descriptions and lists of cited sources. It also recommends site actions involving FAQs, structured data, llms.txt and robots.txt. These suggestions require independent validation: adding a file or schema markup does not itself guarantee inclusion in AI answers.
Shelf connects to MarkApp's Harion, Pantheon and MarkDash advertising stack, turning visibility gaps into media-planning inputs. Earlier ExchangeWire coverage establishes the context of those products but does not validate Shelf's effectiveness. SEO and affiliate teams should assess the representativeness of the question set, reproducibility across runs and any relationship between AI mentions, referral visits and actual conversions. No independent performance results for Shelf were provided in the launch notice.

Комментарии