The Maze: OpenAI is testing a dashboard that would show businesses how often ChatGPT mentions them and how often people click through to their websites, Digiday reports. The tool sits inside a proposed Business Hub, an area for managing how a business appears in ChatGPT. It has no announced public rollout. OpenAI declined to comment, and the reporting doesn't establish who can use the test, in which markets or when.
The Details:
Claim the business first. Ecommerce researcher Juozas Kaziukėnas shared a preview of the claim screen spotted on November 6. It asks users to prove they can represent the business. Kaziukėnas cited connecting a Shopify store as one way to verify ownership. That's an example from the observation; there is no published list of acceptable methods. OpenAI already gets merchant product data through Shopify Catalog so shoppers can discover products. Supplying that data is a separate process from getting access to the test.
Two counts, little context. The reported chart tracks daily appearances in ChatGPT answers and clicks on links to the business's site. It omits the questions users asked, individual user data and competitor rankings. OpenAI hasn't published counting rules for the reported test. For example, it's unclear whether repeated appearances in one conversation would register separately.
A count from OpenAI would fill a gap. Outside services estimate visibility by asking ChatGPT a set of sample questions and checking which brands appear. That can show what ChatGPT says about a brand in a given sample. It cannot show how often the brand appeared in answers to real users across the platform.
MarketMaze Take: If the dashboard becomes available, start by comparing the two counts. Suppose a store shows up more often in ChatGPT answers while clicks stay flat. It has gained visibility within whatever the dashboard measures, but it has no evidence from those counts that the extra mentions sent more shoppers to the site. The store could check its own site analytics to see what visitors from ChatGPT do after they arrive. Testing a few buyer questions might suggest where to look next, though without knowing what users actually asked, it cannot explain the change on its own.
Sources: Digiday · Juozas Kaziukėnas · OpenAI
