2023-02-21
I recently shared why Kiwi6 shut down, mostly focused on its viability as a side project. I thought I would share some more details about why the business didn't succeed.
It's hard to remember now, but the 2010s were sort of a free-for-all for music streaming, and it was unclear who would ultimately win. If we talk about just the most prominent players in the space we had Pandora, Rdio, SoundCloud, Tidal, Grooveshark, and Spotify all duking it out, not to mention iTunes and Youtube. And there were more niche players I admired, like Bandcamp (DRM-free music directly to consumers) and thesixtyone (music discovery and artist tipping platform). But there were clear signs and important reasons to believe that the explosion of platforms was temporary, and that consolidation was inevitable as whoever emerged the leader would use their advantage to build a bigger lead.
One reason is that music platforms have network effects: as a listener, I don't want to have to switch between multiple different apps to hear my favorite music; ideally there's ONE place that has it all. And as an artist, I want to go to where the listeners are. So the "efficient markets" argument is unambiguously pointing towards consolidation.
Second - it's hard to build a popular music service without the most popular music. I feel sort of dumb just writing that. But its true and it has clear and obvious implications. The most important of which is: when music piracy hysteria finally settled down and the major record labels came to embrace streaming, it became increasingly clear that they would be part of music's future, not "disrupted" into irrelevance as many predicted. While they still played a valuable role in the music industry, they also had a singular trump card: monopoly control over the back catalog of top songs that everyone was listening to. And this meant that anyone who wanted to be part of the future of music - both the app platforms and the artists - needed to continue working with record labels, further entrenching their power. In theory the record labels will work with any streaming provider, but in practice you need to be of a certain scale and size to be able to clear the technical, legal, bookkeeping, and financial hurdles to license their back catalog.
Finally, there were the typical returns to scale that aren't music industry specific -- Kiwi6 was often used to commercially pirate music or distribute malware, and attempting to keep this in check required a lot of upfront and ongoing investment that made it difficult to sustain unless you were large enough. Programmatic display advertising was well-established, but online radio ads were still labor-intensive to create and sell -- and building a sales team takes a lot of capital, etc. Music streaming is mostly winner-take-all.
The hard lesson that I (and many others!) learned, very unwillingly, is that there has never been and never will be a large market for specifically "music discovery." Again, the existence of the major record labels explains why. A record label is a one-stop shop for music commercialization, and simultaneously an investment firm at the same time.
A general way to think about how the record label business works is roughly[1]:
A popular but incorrect line of thinking is that record labels don't provide value to popular artists: "After all, they have their own brand, why do they need record labels to promote them?" But this gets the causation backwards - popular artists mostly get popular because record labels took the risk of betting on them when they were obscure [2]. In effect, music labels do the public service of music discovery and curation.
I must emphasize that being a record label is risky, a lot of work, and most of their investments don't pan out. And there's a HUGE imbalance in "who has new music to share" and "who wants to listen to it." Huge! Any platform that wants to make a marketplace for connecting artists and music listeners without a way to address this imbalance is doomed to fail. This imbalance is actually really easy to explain once it's broken down:
It turns out the mass market likes music discovery when its sort of sprinkled in sparingly with the top 40 stuff that's always playing on the radio, and endorsed by someone reputable - that's why live music has opening acts before the headliners. And it pains me to emphasize this: the popular musicians are popular for a reason! If the pitch for your music platform basically boils down to "music, minus all the popular artists" your default outcome is indie musicians begging each other to listen to their own stuff and hoping they can get enough escape velocity to get out of the kiddie pool you've built -- and onto Spotify and a record label deal.
The closest anyone has gotten to succeeding with "music discovery" is SoundCloud, and they're always perpetually on the verge of bankruptcy despite being one of the biggest music platforms AND charging both artists and listeners. And of course - with music streaming being mostly winner-take-all - you don't want to be competing with SoundCloud!
I also missed out on a few trends of the early 2010s, especially the accelerating move to rich mobile apps, as phones continued to supplant dedicated music players -- mostly due to completely predictable incremental improvements to battery life and storage. I grew up…thrifty…so I didn't have a smartphone until like, 2013 (a free Windows Phone I got from work). Needless to say I did not end up with skills in iOS or Android development, and because I was an inconsistent music listener at best, well, I just never connected the dots. Today, people expect to be able to not only listen to music & podcasts on their phones, but to author and edit them as well. So missing out on mobile probably hampered Kiwi6 as a product and business. Whether it would've made a difference in any way is debatable, but "do you have a mobile app?" was never going to go away as a support question.
Podcasting is big business now, and it was a growth area for Kiwi6 that I somewhat reluctantly entered -- we thought our expertise in audio hosting would be highly useful, but were aware of the abstract risks of serving more than one type of customer. Ultimately Kiwi6 was a decent but inconsistent podcast hosting platform, and (as I feared) podcast recording and distribution is only superficially similar to the music business. For one, most active podcasts are regularly updated with new content, and syndication means that listeners don't have to be using the same platform to tune in. For Kiwi6, straddling the competing needs of music and podcast hosting and the contrasting consumption patterns definitely made it hard to succeed and serve the specific needs of podcasters. For example, Kiwi6 still had a storage-oriented pricing scheme, but this was uniquely mismatched with the mental model and value stream of a podcaster - we would charge them incrementally more money over time as they recorded more podcast episodes, even though their experience of the product was that they were receiving a (roughly) constant amount value. Storage costs were coming down and podcasters care less about audio quality than musicians; they very much don't appreciate hitting "storage caps." Hand-waving away the differences between the two industries with "it's just audio" was simplistic and made it harder to win the market.
Finally, one of the things holding back the podcast industry is that its really difficult and labor-intensive to monetize. Because of this there's a catch-22 where only very large and popular podcasts are worth selling ads on, BUT it's hard to get big and popular if you can't generate revenue to invest in your podcast. This is why most podcasts even today (with rare exceptions) are either hobbyist affairs, or essentially a marketing arm of some other business. Kiwi6 might've been able to grow the entire market by providing a solution here, since it was becoming an aggregator of many smaller podcasts. But never escaping subscale and chasing two different rabbits made it hard to focus on this problem. I consider it a lost opportunity to go after an underserved market.
[1] Readers may notice this is suspiciously similar to how the book publishing industry works, which probably explains why the rise of Amazon hasn't led to the extinction of book publishers -- they sit on different parts of the value chain, and specialization/risk tolerance/inertia all provide natural resistence to vertical consolidation.
[2] When artists like Billie Eilish get famous via SoundCloud (instead of being scouted by a major record label before they became big), it's notable that their music careers were launched this way because of how rare it is.