Apps, Mobile & SMS

Roundtable: Is Music Discovery a Feature or Product?

DiscoveryBy Kyle Bylin of

It's an important question. There are hundreds of websites and applications that promise to expose listeners to new music. But can they make money? Are they worth money? This is where the discussion becomes challenging for many music startups. They don't always provide value beyond what established services already do, and in many cases, their music discovery tool is one others could easily build. Thus, it's rarely a defensible business. Below four music and tech experts weigh in on if they think music discovery is a feature or product.

Vote: Feature

Jay Frank | Owner and CEO of DigSin

Music discovery as a whole is best viewed as a feature, not a product. When it’s looked at as a product, it’s likely set up for failure. As a feature, it could produce some great serendipitous experiences. The trick is that these features need to make the discovery seamless. For example, having a created playlist based on your habits that give you new music to listen to without thinking about it. The more passive the experience can be, the higher the likelihood that the discovery process would be effective.

Another important element is figuring out who the music discovery is for. Is it for the listener, the artist or the record companies? Each one would have a different desired outcome to the discovery process. Ultimately it needs to be for the listener, but the listener’s discovery needs are tough to ascertain as factors can be very specific.

What mood are they in that day? How much music do they already know? What time of day are they experiencing the music? Where are they that they are experiencing the music? One song might be a great discovery at 6 PM on a Saturday night at home, that may be an awful discovery at 10 AM on Tuesday at work. How can the discovery feature be able to effectively delineate between those experiences? Can it take into account geographic, age and sex differences between the listening experiences to get to true effective discovery?

Then, isn’t music discovery also about what’s new? Yet, what’s new require data points to be recommended. If it’s unheard, then it has no data points. Then won’t be recommended much creating a self fulfilling prophecy of un-discovery, until something else makes it popular or a manual curation event takes place.

All of this to say that music discovery as a product that people dream of is mostly a pipe dream that is unlikely to be fulfilled. However, a manually overseen product that cross references extensive data points to create a seamless, passive experience might actually have some traction in the marketplace.

Vote: It Depends

Cortney Harding | Partnerships Lead at Soundrop

Eh, that seems a little hair-splitting to me. I get the argument, it’s not really a central point in the conversation. Music discovery services need music to power them, so in that case they are features, but done well they can also be a standalone product.

The real question is how do those features, products, or whatever work, and how does it mimic real life? To go back to my first answer, when I was a kid I learned about Bikini Kill from reading an interview with Nirvana. Then I had to remember them, take a bus to downtown Portland, go to a dark record store (Ozone RIP), and buy the damn record. Now, I can just listen to a track on a streaming service in two seconds, but it still has to come from a trusted source. If I’m a kid and Justin Bieber says “hey, this girl is amazing,” I’m way more likely to check her out than I am to just trust a blogger or a program director.

If I had to vote, I’d vote for curated music discovery over algorithm-driven ones every time. This is more labor intensive, but people can hear and love things that machines never can. I am also super biased because I work for Soundrop, but based on my own experience I listened to playlists on Spotify a heck of a lot more than I ever listened to Pandora.

Vote: Feature

Liv Buli | Resident data journalist at Next Big Sound

In the case of catering to the individual music consumer, I agree that it can be very difficult to monetize music discovery as a stand-alone product, versus a feature that adds value to already existing services such as streaming platforms or digital outlets, or as part of a larger editorial product.

For one, music discovery already exists as a feature across most of these platforms, and while the format, algorithms and underlying data may not be perfected as of yet, the investments we are currently seeing in improving music discovery are great across the board.

Why would your average consumer, provided that they take advantage of these services, invest additional funds in a product that is already being provided?

Vote: It Depends

Jed Carlson | President and Co-Founder at ReverbNation

Passive music discovery is baked into most lean-back music experiences, by definition (the one exception being passive playback of your own music library).

Active music discovery seems to exist as both a product (for example, We Are Hunted’s app and website) and as a feature (social playlists on Spotify). There is also a grey area around things like Spotify apps — are these apps unique ‘products’ or do users view them more like ‘features’ of Spotify? The lines between product and feature get increasingly blurry in a world of platforms and apps. Every day it gets easier to build unique products that feel more and more like native features of the platform upon which they live.

So, in the long run, I think this debate about feature versus product is really just semantics. The question is really around how easy it is for people to become aware of the music discovery tool, how easy and compelling it is to use, and how much benefit one gets from using it. The answers to those questions will likely determine which music discovery tools win in the long run.

Photo by craigmdennis.

_ is founded and edited by Kyle Bylin of Live Nation Labs. If you would like to contribute a post to be featured on the site, please reach out.

Share on:


  1. In the same vein as Jay Frank, and rehashing the obvious, i think we often forget that for the mainstream, casual music consumer, music discovery is not a goal, it’s a mean (to listen to new music that they like.)
    Like Cortney I find the music experience lacking: discovery, listening and next steps should be all integrated into one seamless experience. This has been a problem we tried to address in my first startup (circa 1998.) I also find the discovery aspect either obvious (playlist seeded by something we like) or too random (you might like this) with little way to fine tune it the result.
    I don’t believe there is one “algorithm” that will work for everyone, @ HIVEBEATS we try to emphasize the social aspect with an added layer of source curation, i.e. segregating the people we trust for their music tastes from the rest. We have ways to go and the results are far from being perfect, but we are getting somewhere interesting.

Comments are closed.