UX case study — “Compose your search”
Overview
Spotify is a Swedish audio streaming and media services provider founded on 23 April 2006 by Daniel Ek and Martin Lorentzon. It is the world’s largest music streaming service provider, with over 381 million monthly active users, including 172 million paying subscribers, as of September 2021.
One of the ever-present needs of Spotify is to constantly improve how its users find new music to listen to, thus increasing engagement with the platform and increasing the number of hours consumed on it.
The problem
One of the problems with the platform’s search model that the Spotify research team has found is the difficulty for some users to find music “patterns” that suit their specific musical needs at the time.
Although Spotify has lists of recommendations by categories, some users have expressed frustration at not finding the music that they really want to listen to for the moment or situation in which they are, simply the list of genres does not represent what they would like to listen to at that moment, making the search complex, tedious and sometimes without a satisfactory result.
Goal
The purpose of this exercise is to find a new search model complementary to those currently existing within the platform that allows finding lists more specific to the musical needs of users without necessarily being pigeonholed into a “moment” or “activity”, only based on the music patterns you want to hear.
Users & Audience
Our target is users who do not feel identified with the current search categories within the platform and want more specific results according to their musical needs of the moment.
Design process
Initial thoughts
Analyzing the proposed task and delving further into the problem we find several findings, the most important of which lies in the insight:
“Not everyone understands music in the same way”
While for one person the concept of relaxing music could be classical music, for another person it could be an electronic music list, likewise, for a completely different person the concept of music for training could be reggaeton but for a completely different person it may be heavy metal music.
Having a search from predesigned categories could become frustrating for some of them. On the other hand, when analyzing the direct search by genre, we found a lack of elasticity in finding new musical offerings.
Solution
The design process begins with trying to understand how a search could meet the needs of finding a type of music that suits the user’s particular moment without necessarily using predefined categories or searching by genre.
To tackle this problem we focus on a simple exercise that arises from a question:
What if I can compose my list from musical variables?
The answer to this question could provide a solution to the problem taking into account that:
- The results would not be subject to specific categories or genres, simply to musical variables that can be shared between several tracks of different genres.
- The composition of the search results under this methodology could help to expand the user’s capacity for discovery and musical consumption.
- The search could be easily refined by making the results more precise.
Early sketches
Inspiration
The sketching process begins with benchmarking search systems that could be assimilated to the proposed solution, within the results we focus on the search system used to find new fonts in Google Fonts.
Google fonts is a bank of free fonts for implementation in digital web platforms, to find the right font, users can either enter the name of the font directly or play with a range of typographic parameters that will allow them to find fonts that adapt to the previously consigned criteria.
Taking this into account, we began to develop our own musical variables which the user could modify and thus compose a search query that allows us to bring results according to their needs.
These variables will allow to deliver results closer to the user’s mood, the developed variables are the following:
Lyrics
Determine the intensity of the lyric within the track
Tempo
Determine the speed of the track
Strings
Determine the influence of stringed instruments within the results
Winds
Determine the influence of wind instruments
Percussion
Determine the importance of percussion
Electronic sound
Determines the range of electronic sounds present in the tracks
Based on these variables we designed a two-step search flow:
- Selection of variables
- A single list is presented in the results to start listening to the tracks immediately.
In the sketching process, emphasis was placed on finding a system that allows the set-up of the parameters in an efficient and intuitive way.
Final prototype
For the final prototype, special emphasis was placed on maintaining the simplicity of the solution without losing the aesthetics of the brand in its UI.
Conclusion
This new process allows adding one more search system that responds to the needs raised, making the results agile and assertive for those people who do not want to be pigeonholed into categories and who have specific musical needs that are impossible to meet with the current system.