Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Well, this is another version of the classic "I could build this in an weekend" trope.

9500 is probably excessive, but think of it just this way:

1. Spotify has a worldwide presence, apparently 184 markets; that probably entails a legal presence in many of those jurisdictions, sales, marketing, support, localization, etc.; at a conservative 2 persons per market, just that's going to generate about 400 jobs; now, most of those markets will be grouped up but the big ones will probably have tens if not hundreds of people dedicated to that market (the US is likely to have hundreds of people supporting it, for example).

2. Spotify has to build, update, maintain, extend, etc a global infrastructure. Just the ops team for that has to cover 3x 8h shifts to make it 24/7. Each region should have at least 3 people in it, for high support availability. That's 9 people right there, and it would be crazy to support ~550 million active users with 9 people, they probably have 10 times that many people and teams supporting various components. So just an ops team of 100 is perfectly reasonable.

3. Then they need a dev team. They have... ads, various integrations, songs, audiobooks, podcasts, their apps or whatever are available on smart TVs, web, cars, bla bla bla, if it's 10 people for each client, that would probably mean at least 50 people. And 10 people per client is probably silly low, make that at least 100-200.

4. Then they have a bunch of backend services, probably a lot of them. Plop another 5-10 people for each service. That's going to be many more hundreds of people.

5. Then they have actual R&D, where they're exploring stuff. This depends on the company, but for a company that's still clarifying its business model, having 100 people researching stuff sounds reasonable.

That's ~1500 people just from me eyeballing their business. 9500 is maybe on the high side, but considering their scale, if we dig deep enough into their business model, probably 5000 is perfectly fine.

These services are crazy complex and "streaming mp3s" is a very reductive view.



That, and people often forget that - even if we stay on the engineering side of things - Spotify has tools and apps for artists, creators, and advertisers that most users never see.


I certainly couldn't build that in a weekend but nor would I need 9500 people. Keep in mind WhatsApp was running with 50 engineers in 2015 when it had 900 million users. Last year, according to some sources, it had 3000 employees and 2.4 billion users. And that's groups, voice messages, video and voice calls, WhatsApp for business, integrations and a lot more.

Some of these seem reasonable, some not. There is no need for legal presence everywhere or to even have a permanent hire for that. The many clients make sense (including consoles too) but more than 10 people per client? It's not the most complicated app. R&D I guess makes sense too.

Actually, it would be interesting to see the breakdown of employees by category.

Regardless, I think the answer lies somewhere between what you said (unknown at a glance costs) and what others have said - too many managers managing managers and overhiring.


Whatsapp is just sending messages. Here's music recommendation lecture on using latest techniques to better serve users. I am glad that their music recommendations are very good. There are so many nooks and corners in a big company like Spotify where you need someone to be responsible for.

Music Recommendations at Spotify - Oskar Stål, Spotify https://www.youtube.com/watch?v=2VvM98flwq0


Plus it’s a two sided market - most of that needs to be created and operated for consumer and producer side, and all the systems linking them between.


A friend there says the labels - the producer side - are very demanding to work with. That could be draining resources.


And to factor in some of the other commentary--they could build this in a weekend (and continue to operate it for relative pennies) on-prem.


While I think you're low on the legal/sales/marketing and so on, your estimates does trigger the question: How much easier/cheaper and profitable would it be to run Spotify as a purely stream music platform.

Lose the ads, audiobooks, podcasts, drop any platform that isn't iOS, Android or web. Would that be profitable, or would people not want to buy the service?


Spotify does awesome music recommendation - there are no human curation involved like with Apple Music afaik - you probably need a whole bunch of people for this.

They do much more than merely streaming mp3's and they working on a video product as well - there is some content creation as well i.e agents , sales , marketing and producers.

You can probably can compare them to Netflix (12,800).


I like the way you broke thid down.




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: