Amanda AI, from idea to growth

In this episode of Datadrivet, Torkel Ohman, CTO at Amanda AI, tells the story of how the company went from an idea to a growing business. Amanda AI creates and optimizes advertising campaigns automatically for Google Ads, Meta, and Bing, using machine learning to do the work that would otherwise eat a marketing team’s week.

The idea came from a real problem Torkel hit himself. He had run an e-commerce site with hundreds of thousands of products spread across thousands of categories, a volume that makes building ads by hand impossible. So he built the first version to do it automatically and tested it on Google search ads through a personal drop-shipping site. The results were strong enough to keep going.

The product works in pieces. One module reads a product catalogue and pulls out the details that matter for an ad: the description, image, price, keywords, URL, and stock levels, all into a database. Another turns that information into individual ads tailored for each platform. From there the system manages spend on its own, allocating budget by looking at signals such as impressions, click-through rate, cost per click, and bounce rate. The point is to handle a catalogue too large for any person to advertise manually, and to keep adjusting as the numbers come in. By the time of recording, the company employed 30 people.

The takeaway is a clean example of a product built from a problem the founder actually had. Automation solved a scale problem in e-commerce advertising, and data-driven optimization kept the spend pointed at what was working. That combination is what carried Amanda AI from a single test to real growth.

Listen to the full episode of Datadrivet for Torkel’s full account of how Amanda AI went from idea to growth.