Different platforms personalize for different reasons. I looked at Spotify, LinkedIn, and TikTok to see how they adjust when I interact with content way outside my normal behavior.
On Spotify, I searched for “Bollywood” and listened to music from the first playlist for an hour, even adding about five songs to my Liked Songs playlist. However, despite restarting the app and even checking the next morning, the algorithm’s recommendations on my home page didn’t include a touch of Bollywood. Neither did the daily “Made For You” playlists. I’ve been listening to the same five or so genres for a while now, so I think it’s too used to my listening patterns to shift my taste profile immediately. With more time, I would have listened to a new genre for at least half an hour every day for about a week, and see how long it takes to start registering that as a pattern instead of an anomalous listening session.
On LinkedIn, I searched for “construction” and saved the first fifteen job listings I found, most of them for site manager roles. The job feed responded slightly: two construction jobs showed up in the “Top picks for you” module. But the main feed didn’t shift. It was still full of posts liked or shared by people I follow, mostly about tech and job hunting. The job system responded, but the social content stayed where it was.
On TikTok, I tried a similar approach with two unrelated topics. I searched for knitting, then watched and liked the first ten videos. A knitting video showed up on my For You Page within 13 scrolls. I repeated the test with cooking, this time only watching videos without interacting. It took 32 scrolls to surface one. TikTok seemed much more inclined to rapidly adapt their suggestions based on things I was suddenly showing an interest in compared to Spotify. Since scrolling on TikTok results in me engaging with the content more than looking through the Spotify home page for playlists to listen to, TikTok faces less risk in putting new content in front of me to gauge my reaction.
