A/B Testing Thumbnails and Previews Without Fooling Yourself
Thumbnail tests are easy to run and easy to misread. How to design one whose result you can actually act on.
Articles from Minute.ly on metrics.
Thumbnail tests are easy to run and easy to misread. How to design one whose result you can actually act on.
Most video recommendation still runs on titles, tags and dates — a thin description of a video, and a hard ceiling on discovery.
Most video dashboards are full of numbers nobody can act on. A short list of the metrics that actually change a decision.
Viewers decide whether to click in well under a second, and the thumbnail drives that decision. Most teams still choose theirs at random.
Video revenue is a function of impressions, fill and rate. Most teams push on the one with the least headroom.
Search engines cannot watch your video. Everything they know comes from what surrounds it, and all of that is within your control.