Technology
-
How popularity bias undermines personalization
Recommending best sellers is a good start, but it skews your data toward popular products. Here's how to spot popularity bias and how to get past it.
Read article → -
Multi-armed bandit testing: A/B tests that waste less
A standard A/B test keeps sending traffic to the losing option until it ends. Here's how a multi-armed bandit test cuts that waste, with a worked example.
Read article → -
Ranking products by rating: use the Wilson score
A product with two of two thumbs up shouldn't outrank one with 80 of 100. Here's how to rank by rating with the Wilson score, with a worked example.
Read article → -
How recommendation systems work: a guide for leaders
How product recommendations work: the data they use, how collaborative filtering finds similar customers, and how to recommend new products to new users.
Read article → -
Does your company need AI? How to tell
How to tell whether your company is ready for AI: the signs it will pay off, the signs it's too early, and what to try before you build a model.
Read article → -
How to see what your AI team is really working on
A simple setup that lets executives see their AI team's work as it happens, keeps company data off laptops, and makes the work easy to share.
Read article →




