Spotify (listening time)

Spotify’s personalization investments hinge on listening time. It also has a 30-second rule where a stream only counts if a track has been streamed for at least 30 seconds. This means that Spotify’s advantage in prolonged listening time is on personalized algorithms — helping users stay on the app as long as possible while curating personalization catered to their music taste. Spotify has a lot of options for this, whether that’s the Daily Mixes, daylists, collaborative blends personalized across friends/friend groups, or situational mixes like driving mix, homework mix. It’s clear that Spotify’s ROI depends heavily on personalized recommendations that retain users and convert them to premium.

LinkedIn (session frequency)

LinkedIn’s marketplace hinges on session frequency for users who are job-seeking or job-seekers. This means that LinkedIn has a revenue stream that relies on daily active users and invests in an algorithm that will curate content on the feed and job board that is most relevant for the user. It processes career trajectory on one’s profile and tracks what sorts of job listings are most likely to attract the user to click apply. This algorithm encourages users to consistently return to the platform for more sessions as they see time spent on the platform as a rewarding discovery experience.

TikTok (ad targeting)

TikTok’s personalization system is all about its ForYou page, which is highly curated to a user. A piece of content that’s viewed to completion will surface other similar content that the algorithm has learned about the user. On top of this, the main source of revenue hinges on ad revenue. TikTok relies on ad pop-ups — the more impressions, the better, so the algorithm has to target the right users to increase impressions. These ad targeting have also been sprinkled into and expanded on TikTok shop for a bigger profitable ecosystem.