Enhancing Soundcloud’s Discovery Experience and Positioning through Content Categorization
This is the second part of a case study regarding Soundcloud’s content categorization and exploration of strategic opportunities.
You can find the first part here:
I will be exploring how differentiating the content can:
- Allow a highly tailored experience across different areas of the product
- Help Soundcloud catch strategic opportunities while playing with their own strengths.
It can seem a bit long but there’s actually a lot of screen animations all along!
The roots of a tailored experience: differentiating tracks and mixes
Soundcloud is a content-based product before anything else, just like Instagram, Medium, Pinterest, Spotify, Mixcloud, Unsplash, Youtube etc…
Some of them include one or more types of content, whether it is images, videos, text or audio but one area of focus remains Discovery.
That is delivering an experience that helps users to find more and more content they will love.
On Soundcloud, there’s only one type of media: audio. But if we look closer this audio content can take many forms: track, mixes, podcasts, interviews…
Just like on Spotify you would say! Except for one: mixes. And this particularity can be leveraged to stand out from this competition.
Especially because actors involved in the music production are different from the actors involved in the event production, where most mixes are recorded from.
Now let's see how differentiating tracks and mixes could enable Soundcloud to solve different problems and deliver highly tailored User Experiences in nearly every area of the product.
The Player
Here’s how I revamped the Player screen to leverage those particularities:
- I’ve grouped the “more” and “close” buttons on the top as they are common buttons you can find on different types of screens.
- I’ve grouped all player/content specific buttons at the bottom. I’ve added a settings button that let the user display comments as the track plays and set the audio quality (more on that in “About removing the Go+ tab”).
- I moved and made bigger, the buttons that are specific to the content. More on that below.
- There’s new metadata for tracks and mixes: the record label for tracks and the place where it’s been recorded for mixes. More on those below.
Specific commands
When you listen to a mix, does it makes sense to have a repeat button?
Not really, so that button should only appear when listening to a track.
I also noticed Soundcloud misses a button to see and rearrange the queue list because it lets you add a track to it but you can never actually manage it.
When talking about mixes, none of the above makes sense but there’s something that we can easily observe: users are curious about tracks played in mixes and actually contribute in the comments. That’s an unserved use case that would delight a lot of users if it gets solved.
Here are the benefits it could bring:
- A unique feature you can’t find in any competitors
- Improved discovery experience because you can add tracks to your library for future listening
- Brings a sense of contribution in a community
- Clears the comments section from all these “Track ID?” comments
Improving the library
Spotify and Soundcloud might have different missions, different users and different types of content but they remain music streaming players.
One of the key use cases is to be able to nurture your own library over time but I feel like Spotify is making a better job at this so I inspired myself from their UX to improve Soundcloud’s.
Here are my thoughts about the current layout:
- As discussed in the first part of this story, “Your insights” shouldn’t be there from a listener’s perspective.
- Liked tracks could be treated as a playlist and we could also have “Liked Mixes” and “Liked Podcasts” as we now differentiate the content.
- Albums, EP, Compilation are static playlists created by artists and therefore should have their own menu or tab.
- It requires too many steps to listen to one of my playlists compared to Spotify’s UX.
- “Downloads” don’t make sense here. When I’m offline wouldn’t it better to show which playlists I have downloaded or not similar to Spotify behaviour?
- “Recently played” or “Listening history” shouldn’t belong to my library. I don’t always enjoy what I’m listening to because you give a try to a lot of stuff while browsing Soundcloud. These categories should have their own row in the Homepage.
Leveraging tracks discovery by differentiating listeners, artists, record labels and magazines
Here are, to me, the problems that harm Soundcloud the most:
- Duplicated content when both the artist and magazines/curators/labels upload a track
- Tracks posted by magazines/curators/labels and therefore doesn’t appear on the artist’s page
Which means that when you visit an artist’s page you can’t find all their content, it is spread everywhere.
It makes content inconsistent and hinders the discovery experience.
These problems come from the fact that there’s only one type of account on the platform and that even if the music space has many different actors they are all able to upload content and have the same layout on their page.
A proposed solution
What if artists, record Labels and magazines had their own types of accounts?
Soundcloud already lets artists specify a record label as plain text when they upload new content. But what if they could choose a record label page they’ve been authorized to tag and a magazine/curator to be premiered on?
Think of Medium’s system with publications, authors write their post and can add their post to a publication if they are added as an author.
Once they upload their track, it appears on their page but also appears automatically on the record labels page and the magazine's page. It also automatically posts the content to the feed:
- “Scan7 posted a release”
- “Heist Recordings has a new release”
- “Groove Magazin premiered a new track”
Of course, if you follow all these pages, we should group those updates in the feed to avoid having three different posts for the same content.
The idea here is to only let artists upload content and have it automatically added to related pages through meta-data.
An opportunity for a better user experience
By adopting a tab layout, as in the web version, pages could categorize the content they want to showcase (Releases, Mixes, Premieres, Podcasts) and link-up to other pages they are related to on Soundcloud (Events, Artists, Labels).
To me, this will enhance discoverability because users will be able to navigate endlessly through artists and labels similar to how Instagram lets you navigate from content pages to accounts pages with the possibility to go back to the previous page at any moment.
Another idea is to let artists and labels customize a CTA:
- Labels could have a “Send demo” button to receive and manage demos in a single place instead of managing them by email or messages.
- Artists could have a “Bookings” button to receive well-formatted booking requests.
A bit of user research
As I wanted to better understand how labels used the platform, I reached out to Houseum (45,3k follow. on SC) a well-known house record label and Youtube channel (285k subs.) based in Paris, and I got to speak to Grégoire, a co-founder.
Here’s what he had to say on having different tabs on his page:
“I actually use pinned playlists grouped by year for my Premieres, but I’m currently running out of space for all the playlists I would like to pin. If we could have the ability to have a “Premieres” category grouped by year like I do, it would be a very good idea.”
“I know other media also publish mixtapes, podcast interviews etc…so it might be cool to have other types of categories”
We also discussed letting only artists being able to upload content and be pushed to related pages (labels and media) through meta-data using the mechanism I described above:
“I actually like to take care of the upload on Soundcloud and Youtube myself. The idea is that artists concentrate on making music and then we take care of the rest. Some artists are not really good with communication and digital platforms, they prefer to let us do it.”
“Independant artists should upload the content themselves, but if they have signed on a label, the label should take care of it.”
Good to know and it makes a lot of sense!
That’s how it works on Spotify after all.
This means labels should be able to upload content too but should be asked to set the artist of the release, from a list of artists they’ve been connected with for example.
About playlists
Another big issue is that the concept of playlist belongs to listeners, not creators. Soundcloud has Albums, EPs, Singles and Compilations formats for that but it seems a lot of artists I follow use playlists instead of albums for their releases. I think artists and labels should be able to create playlists from content already on the platform but shouldn’t be able to create playlists from the content they upload.
Leveraging mixes discovery through events and venues
Introducing venues and festival accounts
Now let's imagine venues and festivals could have their own accounts.
That means we could add more properties to mixes like the venue where it’s been recorded and the date. Therefore mixes will automatically appear on the venue page and will be already sorted by date.
Music listeners could then bridge the digital world to the real world and better understand where the music they love is played, discover new clubs, what are their next events and new artists that have played there.
Note: When uploading a mix, artists should also be able to add other DJs to handle B2B cases so that the mix appears on both DJ pages.
Introducing events
As an electronic music promoter, we mainly use Facebook to market our events because that’s where most people are. But I think there’s a big opportunity to catch Facebook’s and Resident Advisor’s use cases by letting users navigate events, listen to the artists included in the line-up and book their tickets from the same app.
Events could then automatically appear on the artist’s pages as described in the previous chapter. I noticed a lot of artists and labels are using the description space to inform about their next events:
Promoters spend an awful lot of money to get their events seen on Facebook because targeting can sometimes be tricky depending on the organization and event size. On top of that, organic reach is so low that you have to pay to be seen by your own fans.
If Soundcloud gets into that game, promoters and venues will have a very good channel to use with the ability to organically reach (through push and email) a more specific and engaged audience.
It’s also an opportunity to generate B2B revenues from promoters and venues by providing a ticketing service which will release the pressure from monetizing from users.
Why not have an event guide?
To push it a bit further, I’ve removed the “Go+” tab (more on that in the next chapter) to introduce an “Event” tab that serves as clubbing guide to let you see listings anywhere in the world.
Improving discovery through the Stream
A discussion about the UI
To start with, I’d like to talk about content layouts in the feed. There are a few things I think should be fixed apart from differentiating tracks and mixes.
To me, the number of plays relates to the track’s activity and therefore is the same kind of information than the number of likes, number of comments and number of reposts. So from a design perspective, those metrics should be grouped.
And I’ll go even further by saying that it’s kind of misleading to have this information out of a cell that is displayed in a feed layout. It lets you think that the likes and comments are related to the post.
On top of that, Soundcloud now lets you repost content by adding a short caption, just like you would on Facebook!
I would not be surprised if they let us comment on posts and reposts in the future and I would personally love it. It would enhance social reactions and we could relate to shared moments with friends.
Here’s how I think feed posts should evolve to bring more clarity regarding what is related to the track and what is related to the post itself.
Getting more control over the feed
In the first part of this case study, I explained why Soundcloud was missing out on the purpose of the feed and social discovery because reposts from friends were drawn into the content published by artists, labels, magazines etc…
A possible solution to this could have been to use an algorithm to give more importance to reposts from friends but in the context of Soundcloud, we want to give everybody the same chance to be noticed. Algorithms are great for feeds to personalize the content but there’s always winners and losers when we use that kind of technology.
So I thought that the best alternative would be to use a filter to be able to choose if we to see only posts from our friends or only from artists, curators, labels etc. In the same way, we could choose if we want to specifically listen to tracks or mixes.
The Home Screen: home of algorithms
When it comes to the home screen, I can see a few improvements to improve discoverability because that’s what the home screen is all about.
Improving social sharing
As explained in my post about Spotify, I think Soundcloud could benefit from having that kind of feature that lets you share content directly to one of your Soundcloud friends. This will avoid having links lost in different messaging service and improve the sense of community on the platform.
Latest content from people you follow
As we now differentiate mixes and tracks, I think it’s pretty straightforward to differentiate “Latest tracks from artists you follow” and “Latest mixes from DJs you follow”.
And as we now differentiate artists from listeners, those lists should only showcase the latest activity from artists.
We also now have magazines account types so we can have a “Latest Premieres” list to go through.
Soundcloud Weekly
Soundcloud Weekly seems to focus on tracks maybe by analyzing a content duration, but algorithms could be even more efficient now that the content is typed.
Another improvement could be to create different recommendation playlists based on different music subgenres like Spotify’s Daily Mixes.
For that, we could make use of one of Soundcloud’s particularity: tags.
But the tagging system should be improved first as tags are not always accurate and don’t always relate to the music genre of a track.
The Dancefloor
A new category could focus on mixes. Same as Soundcloud Weekly, an algorithm could push different mixes based on the user listening history, and what artists and venues he follows.
Track ID?
As Soundcloud has the particularity to be the home of so many DJs and mixes, it could be awesome to have special playlists based on common tracks played by artists we follow.
Gamifying “The Upload”
My experience with “The Upload” is that a lot of the content on this curated playlist doesn’t fit my tastes and that seems logical. If the machine learning algorithm matches you with the content based on who already liked or played it, it doesn’t have much data to work with because it has just been uploaded.
A different approach to this could be to ask producers who want to be featured in “The Upload” to select what are their 10 biggest influences in terms of music and use that to pop their content to listeners who follow these artists.
Above that, I had an idea that could bring a bit of gamification and reinforce Soundcloud’s mission: discovering tomorrow’s talents.
Let’s be honest, when on Soundcloud we are more likely to go and listen to content pushed by an algorithm based on our listening history than content with only a few plays. Also, our brain has a bias that led us to think that a track or mix is better because it has more likes or plays.
Imagine if every day you were given 10 new uploaded tracks in “The Upload” but the artist, title, number of plays, reposts, comments and likes were hidden.
You then have a certain amount of new virtual resources to distribute to tracks you really liked.
If a track reaches a certain amount of that virtual resource, the track is revealed to you.
This concept is based on unpredictability and variable rewards which are important Gamification mechanisms I’ve been studying through Nir Eyal’s Hooked model and Yu-kai Chou’s Octalysis framework. The concept leans on the “rewards of the tribe” and “rewards of the self” in some ways.
Of course, this is just an idea, but I think Soundcloud should work on those kinds of initiatives to reaffirm their position and what they stand for: revealing artists through their community.
About removing the “Go+” tab
As discussed in the first part of this case study, I’d remove the Go+ subscription to have only the Go plan which includes: no ads, better audio quality and offline mode.
Put simply, I think Soundcloud is a bit too begging with monetization. It harms the experience and brand image.
The only strategy I can see in that approach is they hope the user will finally hit the “Try Go+” button when stumbling upon an ad once again. Having said that, I don’t work at Soundcloud and this might actually work well.
On the other hand, the player screen when an ad is played is poor in trying to convert users to Soundcloud Go: the user flow is not optimized, and branding is nearly inexistent.
If it hasn’t been done yet, I would probably invest a bit more in experimenting on different flows to see how they convert users to subscribe to Soundcloud Go.
Here are some of the best practices in monetization I’ve been learning about and experiencing through Madly (the dating app I’ve co-founded):
- Make your Premium features part of the UX to make it clear on what you’re offering.
- Make it easy to subscribe by optimizing user flows.
- Copy-writing is king. Personalize it regarding the context.
- Trust is one of the drives of e-commerce and trust happens through delightful UX and branding.
Here’s how I would showcase the ad-free and better audio quality value propositions of Soundcloud Go. These are only starting points and other approaches should be A/B tested to see how it impacts metrics like conversion rate but also retention.
Here are a few ideas I have to make Soundcloud Go more appealing while playing with Soundcloud’s own strengths:
- Following Spotify and Netflix’ success, offer shared plans for friends: Soundcloud is more social than any of these products, I can’t see why it wouldn’t work.
- Ability to see and contribute to mixes tracklists.
- If adding a ticketing service works out, offer a small discount on tickets or pre-sale access for subscribers to nurture the sense of community.
Taking a step back
If you are not from the electronic music space you must think I have taken some of those ideas very far!
And that’s the main problem when you’re passionate about a product and sit among the power users, you tend to forget you only see the product from your own reality.
Therefore those changes should only be considered after massive user research and going deep into Soundcloud’s data to validate the problems and solutions I’m highlighting, how much users would be impacted and if such moves won’t impact negatively other user segments.
Here are a few of my persona characteristics that led me to design those changes:
- Listen to music daily, +5 hours
- Listens almost exclusively to electronic music (deep, house, disco, techno, minimal-house)
- Is curious about new stuff coming out by following closely artists and labels
- Going to events/clubs is part of his lifestyle (before Coronavirus 😥)
- Promotes electronic music through events, booking local talents and international artists
- Has 7+ friends on Soundcloud that reposts music they discover
- Shares music with friends weekly
This leaves me very curious with a lot of questions about Soundcloud’s user base:
- How do I fit in Soundcloud’s user base?
- What is the part of electronic music listeners?
- What percentage of active users have 7+ friends reposting content?
- What is the importance of events, labels and magazines in other music genres?
- What percentage of users listen to other types of audio (politics, spirituality, economics etc…) and what is their LTV?
So I decided to reach out to friends who are into different music genres to know a bit more about how they use the platform and the importance of events and labels in their eco-system.
Pierre, which has his own Rock/Stoner group called 7 Pills:
“Soundcloud is not very popular in the Rock/Stoner community. We mainly use Spotify or Bandcamp to discover new bands”
“Of course labels plays a big role in the eco-system but there’s not as much events as in electronic music so I don’t think this genre will benefit so much from having events and venues on Soundcloud.”
Luca, sound engineer in a Studio and works mainly with rappers:
“Soundcloud is not very popular in the rap space because their tracks often come out with a video clip and it’s the way this music is consumed. Rappers mainly diffuse their creations on Youtube or even Instagram TV. Some of them still have Soundcloud accounts but their number of followers is nothing like the audience they have on Youtube and Insta”
“There’s quite a fair amount of clubs that plays rap music tho and there’s DJs as well. I don’t know if it’s very common that they have a Soundcloud account and an audience there.”
I’ve also searched for the most popular french independent labels in Pop/Rock/Folk and searched for their account on Soundcloud. Out of 12 labels on that list, 8 have a Soundcloud page with a fair amount of followers: Musique pour la danse (21k), Born Bad Records (8,6k), Pan European (8k), Hands in the Dark (4,1k), Howlin Banana (1,9k), Objet disque (729), Jeune à jamais (94), Buddy records (33). I noticed they only upload tracks but you can find some interviews on the web, so showing them they can showcase different content could inspire them to add more to the platform.
I think the importance of labels is a common denominator across all music genres and empowering them will benefit the whole platform.
Introducing events and venues will only resonate to particular music genres but I can’t see how it could have a negative impact.
Time to sum it up!
If I had to concisely express a vision to align different teams to work together, that would be:
“Empowering every actor of the music industry to showcase their work and create relationships in the digital space.”
This project has been an incredible journey and I had so much fun going deep into one of the products I love the most.
It was also my first big project (50 screens) on Figma and enabled me to up-skill a lot! I seriously think I will never go back to Sketch again.
Here’s is the full prototype if you want to have fun exploring:
Prioritization
I believe that Soundcloud’s biggest problems today are:
- Monetization on the user side, that is converting more users to Soundcloud Go.
- Losing users to Spotify because of a more consistent catalogue.
To me, both of those problems rely on the same root problem: duplicated and inconsistent content.
Why would I pay for a service that doesn’t have all the content from artists I follow when it’s actually available on Spotify?
So the first thing I would fix is differentiating listeners, artists, record labels and magazines. Then create a system that only allows artists and labels to upload content with the correct metadata for it to be added to artists’, record labels’ and magazines’ pages (for Premieres).
This is the most challenging task as it asks to update existing data that have been there for 14 years and will probably need to ask users if they want to convert their account to an artist/label/magazine account.
Ideas are nothing without execution
In the coming weeks, I will continue to work on that project by thinking of the different initiatives needed to validate those assumptions and building a prioritized roadmap (probably on Productboard) taking into account technical challenges and operating those changes incrementally and smoothly toward users. I’ll also think about how we could organize a product team to support those changes.
Feedback very well welcome !
Thank you for reading!
If you liked what you’ve read please fall asleep on the clap button and leave a comment with your thoughts, especially if you’re into a different music genre!
