40 % von SerpAPIs Umsatz stammen von AI-Chatbots

Ein kleines Follow-up zu meinem Artikel über Googles Suchdaten vom Mittwoch, nachdem The Information in ihrem AI Agenda-Newsletter weitere Informationen über SerpAPI geteilt hat:

Around 40% of SerpApi’s business comes from such AI startups, according to CEO and founder Julien Khaleghy. Aside from OpenAI, they include Cursor, Perplexity and Meta Platforms. Another 40% comes from marketers doing search engine optimization, and the remaining business comes from more creative applications, like banks doing background checks on customers, he said.

That’s not how SerpApi started. Khaleghy told me that he founded the company in 2017 when he was trying to find a faster way to mass download images of food from Google Images so he could train an AI model for a calorie counting app. At first, most of the company’s business came from marketers who wanted to see how well their companies were ranked on Google Search, Khaleghy said.

But over time, more AI startups became interested in SerpApi’s services to either mass scrape the web for training data or to do real-time searches to provide their chatbots and models with up-to-date information, he said.

Man habe außerdem keine Angst vor einer richterlich erzwungenen Google Search API, weil Google diese womöglich absichtlich langsam oder schwer zugänglich gestalten würde. SerpAPI arbeitet indessen an einem von Google unabhängigen Suchindex und will in den nächsten Jahren an die Börse gehen.