I could write Spotify in like a weekend or two tops! I don't see the need for all that dead weight. Personally, I don't like coding mobile apps, so idk maybe hire like 2 people (1 Android and 1 iOS) dev for that. I'll do the backend. Easy savings really.
> The hard part of spotify is getting the deals and the labels and the users.
Having worked for a (now defunct) company that was trying to get labels onboard, this is exactly correct. They had 3 or 4 people (plus the founders) who mostly spent all day chasing up music industry contacts and lawyers trying to get them into conversations and persuading them to license their content. And having had to deal with CWRs and ERNs, that's also a nightmare of hideous proportions you have to deal with for streaming services.
I can't even tell if this is parody. I, too, can write an app in a weekend that streams nothing because I own no servers, no disks, and have no licensing agreements with rights holders allowing me to host and serve any of the content they own. Surely, writing the app is all there is to this business.
For normal tech companies with low marginal costs sure. Spotify pays 70% of revenue straight to labels though, they need far more revenue per employee than normal tech companies to make the math work.
It's an interesting question. I guess they are mostly sales or support people, which need to be scaled linearly to a number of customers, labels, artists, etc.
I guess people misinterpreted the parent's statement. Spotify had 9,000 employees and downsized to 7,500. I think downsizing to 3,000 is a quite reasonable ask.
As the owner of the site previously known as twitter showed - once you are out of the hockey stick phase you could do with a lot less than anyone thought possible.
They had thousands of engineers alone. Many years ago, Spotify adopted the infamous "squad" model where each individual UI element or feature would have a dedicated team of 6-8 developers, designers, and PMs. Each of these teams would be organized into higher level groups with yet more cooperation staff, and so-on for a couple more levels. Add on all the people making tools for developers, admin, the thousands of additional people to deal with the music industry/sales and yeah, not hard to see where the headcount came from.
They changed structure a couple years back, but headcount didn't really go down.
According to their career page, DocuSign has 77 open roles in engineering. That's feels like a lot, but given they have identify verification where each country has nuances, maybe there's a lot going on behind the scenes.
This feels like a clickbait title. I don't think that Spotify is in that unique of a position compared to other tech companies that have gone through "hypergrowth" and are now focusing more on profitability.
I said this a ways back in another thread, but I think over the next 10 years or so, we’re going to find out exactly how much of the tech sector was built on the back of absurdly low interest rates making money basically free. Just borrowing and borrowing and borrowing over and over gesturing vaguely to the concept of “behavioral data” in the ethereal future and how you’ll make it all back advertising mattresses and food subscription boxes.
I read the whole article but didn't learn anything new or interesting. Why Spotify is screwed? I thought the music labels have leverage to squeeze Spotify or something like that but the entire article is just stating what has happened with the recent layoffs and financial statements
IMO there would be absolutely nothing wrong with going back to a pre-streaming world. I use Qobuz and Bandcamp to buy albums with no DRM already, it's great. It's more expensive than using Spotify, sure, but that makes sense, as the article makes clear. Bandcamp is proof that more equitable business models are possible.
Had a convo with a top Spotify developer getting paid really well. He was, I thought, incredibly arrogant, he talked about Spotify as if was OpenAI or something.
He was open in discussing how they only want the best candidates and how hard their interviewing process is, he explained that is why they are so successful.
I pointed out that Spotify is losing money at an unsustainable rate. The whole thing to me was comical. Spotify exists in a space that is a complete commodity at this point. I switched to YT music recently (because I didn't want to see ads) and to me YT music is as good or better.
Spotify had a sort of first mover advantage but drastically over hired like Twitter and everyone else did. I guess we'll see if they can get to profitability.
I guess you mean the auto-playlist/radio stations play the same 20 songs?
I've had same experience as parent, Apple/Spotify/YT Music are interchangeable but for a few exceptions in who got a license to what, in which case YouTube wins because of bootleg uploads (Mos Def's Ecstatic comes to mind, ain't nobody got a license to stream that, but there it is on Youtube)
For my taste Yt music has almost everything, but spotify hardly has half my playlist.
I've found most of the music I listen to through youtube. So it's not suprising.
But I'd wager that almost every song on spotify is uploaded to youtube as a video. Then it should be available on yt music unless they have flagged massive amounts of videos as not musical.
> how they only want the best candidates and how hard their interviewing process is
I sometimes feel like those algorithmic challenges at interviews are more for making developers feel they work at the "top team" than filtering the candidates. I am not salty about it - I am very good at those, did lots of Topcoder back in the day.
Most of the people in the industry agree that they have a terrible signal-to-noise ratio (or any other standardized interview method for developers). Any method that works doesn't scale, but large IT corps don't want to hear that.
> I sometimes feel like those algorithmic challenges at interviews are more for making developers feel they work at the "top team" than filtering the candidates.
I think that's partially correct, but a side effect. I think the real function is to make managers and execs feel that their devs are the "top team". But totally right about it being more about feelings of superiority than objective and relevant qualities for the job.
Makes sense there is a lot of self worth or pride people attach to their work. It is quite common in software devs I know.
Especially young and impressionable ones. I am past 35 so I am past all of that. I know I do good work and I don’t need external validation as much as I was 20-something.
So long as those algorithmic challenges are actually in-use. Warranted. If it's just to flex that last little bit of grey matter, not warranted. The issue is software complexity is such a spectrum that so too must the interview process. No standard "developer 101" interview exists because each company has different software needs. An entry-level exam at Google will look different than an entry-level exam at Dunder-Mifflin. I think better expectations and better communication are the way to have successful (and meaningful) interviews. Yes, code test and all that but you should formulate the interview so that it gives you a good understanding of what they would be like to work along-side. I agree though with what you said.
> Spotify exists in a space that is a complete commodity at this point
There is massive lock in since they have years of my music listening history and have the best recommendations around. I also share an account with 2 other people in the same boat where it'd be extremely disruptive to switch.
That's hardly a commodity type situation and what you listened to + how long per song has never been a transferable thing in the industry and music recommendation algorithms significantly fluctuate in quality (youtube only redently improving theirs thanks to Tiktok competition doing it better than them). Spotify is top of the field. The hifi (finger quotes) stuff like MQA they have coming soon means they'll be competitive with TIDAL for quality, and there's only 2 hifi options in the industry with large libraries.
More anecdata: I do not mind much my spotify historie. We choose mainly radio channels or listen to a specific artist. In general (news, music, movies/netflix!) I think recommendations suck big time and are overrated. Better ask friends.
No way I'm an exception. Every person I know is obsessed with the weekly discovery list... It's a major reason Spotify is as popular as it is.
It sounds like you don't really listen to music if you're only getting the same songs. I don't see how that's possible in any other way, unless youre the one listening to the same small group of songs and it has no data to make useful recommendedations.
Weird it's a common thing that music discovery was basically solved by Spotify. You don't have to follow blogs, magazines, review sites, or seek out recommendations etc. Making playlist for friends was also common but now I feel way less of a need now since I know Spotify will do a better job of tailoring it to them so I just occasionally share individual songs.
I'm 35 not sure how old you are but Ive been using Spotify forever and recommendations has been a major part of that UX. And I used Last.fm before that like 40 million other people, which Spotify killed after ending their deal with lastfm and building their own one.
Considering you don't get proper recommendations says enough about how much you use it IMO. You won't get anything back if you don't tell it what you like aka listening to music often.
I'm 42. I listen to spotify nearly everyday, though I'm not nearly as interested as you in finding the latest music as you are.
I'm sure there is a significant segment who loves the discovery like you, and that segment is probably in the millions buy its probably still the minority.
Even in this thread you are the only one I've seen mention this specific aspect.
Most people probably don't care except for a segment of ultra interested people like yourself.
Fwiw YT Music gives me a huge variety of songs it suggests for me at least.
A good example is radio. If you pick a song and do radio Spotify will start to repeat songs after 7 or 8 of them. YT doesn't do it nearly as often.
I use both and I prefer the Spotify interface by a mile, but I'm "stuck" with YT Music because they let me upload my own MP3s (Spotify does not). Without that ability, I can't make playlists I want because I don't have access to the songs I want.
But yeah, they bare essentially interchangeable, unless you care a lot about your "year in review" stuff.
It’s extra sad, because you used to be able to put your own MP3 files in Spotify. If I recall correctly, it would even sync those files to other devices!
You aren't mistaken. I used to use Spotify for that purpose as well.
I had an account with Spotify premium, which allowed you to download playlists to listen to offline. On my computer, I would select a local song to add to a playlist, and on my phone I downloaded it to listen to.
I believe it still does that, but cannot confirm, because the way it syncs the stuff is kind of aligning the stars in a specific way. I've followed instructions and it never seemed to work quite right. It skips some, it dupes some, some might only work on the same network (even when everything is marked for offline)
I do have local stuff (mp3 based) on my Spotify still on my iPhone. Not sure if it is a relic of an old feature. But it's not the entire directory and it's not easy to manage.
They aren't screwed, since apparently you can't make any changes to your account plan. I've been trying to downgrade from family to individual for a week and I keep getting an "oops...try again later" error.
You can't lose business when you don't let your customers change their plan!
Speaking purely from the dev side, a competent 5-person team could probably build 95% of Spotify’s current functionality in 2-3 years. Of course there are many other necessary nontechnical roles that would be unfilled (legal, business development, etc).
It’s no secret that companies tend to scale their workforce exponentially in exchange for logarithmic gains in their core product.
That's quite the leap you made there to get dramatic about LLMs takin' our jerbs. It seems to be a trend...
Support is not likely it, it's not the type of app that requires a lot of human support in the first place, tbh, I'd be surprised if you can even talk to a live person. Even if it was support, those would probably be the lowest-paid employees and not help much in cost cutting.
Not sure what a LLM would even provide for Spotify support that a static FAQ couldn't.
From my experience this is wildly incorrect. Not even big tech hires that many devs, especially for frontend. That's at least half the reason why most frontends are terrible.
Spotify had roughly 9,000 employees prior to the layoffs. What if they took the approach of X and let go of 80% of their staff - where would that put them?
1,800 staffers.
Just throwing this out there - I've worked for companies producing far more sophisticated products with under 500 employees in total. I'm honestly not understanding why you would need 1,800 employees - and that's an 80% reduction from where they were!
reply