An algorithm is a set of instructions and multivariate calculations that learn and adapt over time. The premise for using algorithms for digital advertising are enticing because an algorithm can take in bigger data sets than any person could analyze on their own and make smart decisions based on the instructions it is given. An algorithm should meet customers’ goals and create value in the most cost-effective and transparent ways possible.

Unfortunately, the word algorithm is often one of the first talking points that vendors in digital advertising refer to but can say very little about given that most platforms are built on proprietary technology.  This overuse of the word has created ‘algorithm fatigue’ and resulted in loss of meaning. The value proposition for many vendors is, “our algorithm does the heavy lifting and will meet your advertising goals, so sit back and don’t worry about how.”

The problem is that I want to worry about how. I’m glad that the algorithm has figured out how to meet my advertising goals, but were they the right goals- should I have set them higher/lower? What does the algorithm know about my customer, competition or industry that I don’t? How much money is it keeping for itself from arbitrage between me and the publisher?

Vendors and advertisers alike should focus the conversation around algorithms on customization to fit their needs and bottom line revenue results that are transparent.