This post was originally published in March of 2017 on another platform:
One of the selling points of the NPR One app is that it follows what you listen to, then makes suggestions about other things you might also be interested in based on your tastes. It learns your tastes by noting what you listen to and what you don't listen to (skip). This pattern of recommending may sound familiar. If you've ever ordered something from Amazon you will recognize the suggested list that says something like "other people who ordered what you did have also ordered these…" Even more recently I noticed that Amazon noticed what I looked at but didn't order. After logging on I got a note that said something like "based on your recent searches you might be interested in some of these related items."
Here's yet another story in IEEE Spectrum of how Spotify is jumping on the curation-suggestion-individualization bandwagon:
http://spectrum.ieee.org/view-from-the-valley/computing/software/the-little-hack-that-could-the-story-of-spotifys-discover-weekly-recommendation-engine
In this case, the idea/project was started by some engineers within Spotify. The recommendation tool at first didn't take off. One of the creators shared, "My hunch was that navigating to this page and looking at albums was too much work." The original tool required customers to go and check out the suggested content. Gradually they developed the more proactive tool. The article shares, "Their system looks at what the user is already listening to, and then finds connections between those songs and artists, and other songs and artists, crawling through user activity logs, playlists of other users, general news from around the web, and spectragrams of audio. It then filters the recommendations to eliminate music the user has already heard, and sends the individualized playlist to the user." Without telling people, they pushed out the feature to Spotify employees. Reaction was positive. As the tool become popular internally, Spotify decided to put it into the production system for customers.
Whether you think this sort of thing is helpful or creepy, it's clear that companies believe it adds value. I'm not sure there is a place for this particular idea in all applications, but what I find interesting is that the idea came from someone seeing a need and a solution without waiting for "management" to point them down a path. From the article, "'This wasn't a big company initiative,' Newett says, 'just a team of passionate engineers who went about solving a problem we saw with the technology we had.'"
One of the selling points of the NPR One app is that it follows what you listen to, then makes suggestions about other things you might also be interested in based on your tastes. It learns your tastes by noting what you listen to and what you don't listen to (skip). This pattern of recommending may sound familiar. If you've ever ordered something from Amazon you will recognize the suggested list that says something like "other people who ordered what you did have also ordered these…" Even more recently I noticed that Amazon noticed what I looked at but didn't order. After logging on I got a note that said something like "based on your recent searches you might be interested in some of these related items."
Here's yet another story in IEEE Spectrum of how Spotify is jumping on the curation-suggestion-individualization bandwagon:
http://spectrum.ieee.org/view-from-the-valley/computing/software/the-little-hack-that-could-the-story-of-spotifys-discover-weekly-recommendation-engine
In this case, the idea/project was started by some engineers within Spotify. The recommendation tool at first didn't take off. One of the creators shared, "My hunch was that navigating to this page and looking at albums was too much work." The original tool required customers to go and check out the suggested content. Gradually they developed the more proactive tool. The article shares, "Their system looks at what the user is already listening to, and then finds connections between those songs and artists, and other songs and artists, crawling through user activity logs, playlists of other users, general news from around the web, and spectragrams of audio. It then filters the recommendations to eliminate music the user has already heard, and sends the individualized playlist to the user." Without telling people, they pushed out the feature to Spotify employees. Reaction was positive. As the tool become popular internally, Spotify decided to put it into the production system for customers.
Whether you think this sort of thing is helpful or creepy, it's clear that companies believe it adds value. I'm not sure there is a place for this particular idea in all applications, but what I find interesting is that the idea came from someone seeing a need and a solution without waiting for "management" to point them down a path. From the article, "'This wasn't a big company initiative,' Newett says, 'just a team of passionate engineers who went about solving a problem we saw with the technology we had.'"
RSS Feed