How AI Search Changes What Customers Find — Kartik Hosanagar

Kartik Hosanagar makes a distinction that matters for anyone who wants to be found. Search was a contest for position: a buyer scanned a page of links and chose among them, so the work was to rank higher than the alternatives. When that same buyer asks an AI assistant instead, they receive one answer and act on it. The work is now to be the answer. Hosanagar frames these as the same goal pursued in a completely different sport.

The practical consequence is that the established discipline does not transfer. Ranking was measurable and its levers were reasonably well understood. Being recommended by a model is neither, and Hosanagar's point is that no playbook exists yet. Tools have appeared that score how often a brand is recommended in AI answers, which measures the outcome without explaining how to move it.

For companies this becomes a question about how narrow the path to a customer has become. A results page offered options and let the buyer choose among them. An assistant offers a conclusion. That raises the cost of being absent from the answer, and it puts a premium on the things a model can actually read: clear, factual, well-structured information about what a company is and does. Hosanagar has spent two decades researching how algorithms shape what people see and buy, and co-founded Bodhium Labs to work on this problem directly.