How to make Shazam a more social app
Add a Feature: Shazam Party Mode
The brief
The brief for this first individual project was to add a feature to Shazam that would enable people to share their playlist. I wanted to practice a bit more with coming up with new ideas for apps, so I decided to broaden my search for a new feature using a more general brief: how to make Shazam more social?
Research
I started, of course, with research. I found some specific information about Shazam users online, and it turns out they are mostly male, between the ages of 16 to 34, and have a relatively high income. I also ran a quick user survey and found out most Shazam users use the app once a month on average, and mostly at bars/clubs or in shops and cafés. One of the most interesting things was that most people don’t actually use the existing social feature where you can share a song you have Shazamed with others (by sending a link). When asked why, most people said they didn’t know this feature existed. But now that they know about it, they still weren’t very interested in using it. It turns out that Shazam is more of an individual thing: people use it to learn which song they are listening to, but don’t really feel the need to share that with others.
Generally speaking, people do really like to share their musical taste with others though. People like to tell other people about new music they have discovered, and people like to get recommendations from other people (or other apps, such as Spotify) about which music they should be listening to. So how to incorporate this into Shazam? How to make an individually used app more social? Those were going to be my main questions for the brainstorm and concept testing.
Brainstorm & Concept
As this was an individual project I thought it was a lot harder to do a proper brainstorm. I did try to come up with some ideas on my own, but I ended up spending a lot of time staring at a blank piece of paper. So I decided to ask a couple of friends and I brought a bottle of wine to get the ideas flowing. And once they got started, they didn’t stop, it was hard to keep up with all the ideas, actually! I ended up with 6 different concepts that I thought sounded like feasible ideas (or just really fun), so how to choose between them? I decided to do a quick user survey and have people vote on the concepts.
The winning concept was something called “Party Playlist Mode”: being able to influence a playlist on Spotify that is being played at a party, using Shazam.
Prototyping & Testing
The next step was to come up with a way to implement this concept into the existing app. As soon as I started drawing my lo-fi prototype I realized I had a couple of decisions to make. First of all, how does Shazam know which playlist to influence? While making my site map and user flows for the original app I had found out that Shazam tracks your location, so I decided to use that to connect to specific Spotify playlists. You will be able to influence any playlists that are playing in your immediate vicinity, so the host of the party can just turn on their favourite party playlist on Spotify, and then the guests can use Shazam to influence that playlist, to make sure everyone is hearing something they like.
And then, how to give people the power to influence the playlist? You don’t want them to fiddle around too much with their phone during the party — people should be dancing and having a good time, not staring at their phones — so I decided to use something very quick & easy: upvoting and downvoting on the current song. I would then let Spotify decide which songs to play next, using its own algorithm. So for example, if people like the song that is playing and upvote it, Spotify will search for songs that are similar and will add them to the playlist.
I tested my assumptions with the lo-fi prototype and got some good feedback on that. Using the feature itself was not very complicated, just using a toggle to turn on “Party Mode”, and then voting. The process was easy to understand and didn’t require a lot of adjustments. But my tester did have some good questions: where is the flow for the host, how do they turn on Party Mode and select the playlist? Can we have more general options to select, like how many beats per minute I want the songs to be that Spotify chooses next? Can we request specific songs?
In the next stage, the hi-fi prototype, I didn’t change that much about how the feature was implemented in the original app, I just changed the look of the Party Mode button and voting icons to make it more in line with Shazam’s general UI. I also added a user flow for party hosts, and some onboarding instructions. I then went back to testing and ran my prototype in Maze.
Again, the users didn’t have a lot of problems with turning Party Mode on and off, though I did get some feedback on the onboarding instructions being a little too slow and sometimes confusing. The most interesting results from this second round of testing was that people were more likely to use the app to influence a playlist as a guest (4.1 out of 5) than using it as a party host to give guests the ability to influence the playlist (3.8 out of 5). From the comments made by users I realised they were worried people would abuse the possibility of influencing the playlists, by choosing music that didn’t fit the party mood or songs that the hosts didn’t like at all.
I made a few changes to update the hi-fi prototype, based on the user test results: I removed a lot of the instructions except the pop up in the first screen, and gave people the possibility to close the instruction window, to make it less cluttered. I also reduced the number of host screens to make the process more straightforward. Based on some of the user feedback I also got started on two additions to the Party Mode feature: having the playlists in your Shazam library and the ability to request specific songs, but I decided both didn’t actually fit that well with the original concept. The playlists are on Spotify, so I think it would just be more confusing to have them in your Shazam library as well, especially as Shazam also has its own generated playlists there. And having the ability to request specific songs, well, in the end I decided it would still be too fiddly during a party. I realised that, despite several users requesting this feature, I hadn’t actually changed my mind about this. I think it would be more useful at this stage to focus on explaining the original concept to the users instead of adding a feature that changes the concept.
The full version of my latest hi-fi prototype can be found here.
Next steps
So what’s next? Obviously, as mentioned before, some work needs to be done on explaining the concept and why it’s kept simple. I would like to create a few possible options for the onboarding instructions and then A/B test those on possible users, to make sure the instructions are clear and useful. Also I would like to test the concept a bit more, to see if I am right about requesting specific songs being too fiddly during a party. And lastly, as some users commented they would worry about who was voting on the songs and possible abuse of the option, I would like to add some security measures. Right now it’s just based on location, so anyone in the vicinity of the host could influence the playlist, but it might be useful to give the host some options to kick specific people off the playlist, or have it work by invite only. So that is something to explore and test.
Once these kinks have been ironed out and I have a well functioning MVP, I would like to start adding a few things. I think this concept would be very nice for cafes and shops to start using to give their guests some influence on the playlist, so I would like to add some screens that cater specifically to cafe or shop owners. It might also be interesting for DJ’s to use the concept, but I imagine they would want to have a lot more influence on the process and not give people that much control over their playlist. So more research will have to be done on if and how DJ’s would like to use Shazam to get feedback on their playlists.
Learnings
I really enjoyed doing an individual project as it meant a change of pace and the opportunity to stick to my own schedule. It was also very interesting to see which parts of the design process I’ve got covered, and where I still need some work. For example, I’d never actually written the User Persona before, so it was good practice to do that.
But I did end up really missing the creative input from other people. You really need a couple of people to get together to have a proper brainstorm session and to feed off of each other’s ideas. You get a lot more creative when you’re with a team than on your own. Whenever I would get stuck in designing the prototypes it would’ve been really good to have someone to bounce ideas off. A group of friends and a bottle of wine can get you quite far and leads to some surprising insights, but sometimes you really want another designer to talk to!
