- daily updates is gone. This is where I can get a snapshot of all new releases from artists I love. I’m not sure if Release Calendar is the new one but I don’t bother to check.

- I listen to classical and the song title naming is just subpar. For example, “Well-tempered Clavier, Book 1, J.S Bach, Angela Hewitt, Prelude in C Major is too long to be in the title. Especially, the rat if the album is just a repeat alternate between prelude or fugue and the chord.

- recently play list stopped syncing between my phone and my desktop app after 2022 for some reason. Is it a bug or they just stop doing this since it costs more to sync?

I probably won’t switch to a different stream service for now as there is not much differences for me to migrate.

Even assuming delivering music, podcasts, royalty payments, negotiations with artists, etc..., takes 100x more headcount in total, that should still be only around 2000 employees.

9000 just means most are not being that productive.

Can you link the source?

But I wonder if they have to replicate non-dev teams in multiple countries?

I’d imagine they have to a seperate legal, promotional, marketing, sales and so on teams for each market.

Replicating these business functions quickly adds up to thousands of people employed!

As an English speaker I tend to think of music as being all in English, but each region has its own musical culture that would need addressing.

France I believe mandates a certain amount of French speaking songs.

Spotify has a completely different business model, architecture etc.

In the total category, R&D (1,257), S&M (1,101), G&A (430). R&D trending up, possibly plateauing, S&M, G&A trending down. [1, pg4] All relatively minor compared to 7,159. (

[1] pg 36, Spotify 2023 Financial Report, "Spotify Technology S.A.", FORM 6-K, October, 2023 - https://s29.q4cdn.com/175625835/files/doc_financials/2023/q3...

The R&D expense is fully within their control, while things are a bit more complex with licensing costs.

i’ve used her phone to control music on long car trips, if there are differences i haven’t noticed. again, anecdotal.

I restarted my phone, reinstalled the app, went through several iterations of iOS updating, the problem never went away. No other app, including Spotify, has this issue on my phone. After a month of paying for Apple Music without Classical working I shrugged and unsubscribed. Big loss for me, and for them, since I am exactly their target audience with Apple Classical.

https://support.apple.com/contact

https://discussions.apple.com/thread/250348779

I thought this was just me. Defo a bug.

Having had a wee insight into how it worked at Rdio (obligatory: I miss Rdio!), one big chunk of the answer is they need a literal army of lawyers.

A worldwide streaming music library is a hoard of licensing liabilities for every possible country, with deals that expire, renew and change hands daily on every possible mix of business days and holiday calendars.

As for the design and development headcount, I guess they dispatch teams and tasks as efficiently as Meta with their thousands of acqui-hired product people.

As much as I hate to see layoffs from an empathization perspective, maybe this is the reason? Remember Brook's Law - adding manpower to a late project makes it later. Too many cooks spoil the broth and all that.

1. iOS App 2. Android App 3. Windows App 4. Microsoft App 5. Web App 6. Underlying API 7. Artist Portal/App 8. Advertising Portal/App 9. PlayStation App 10. Xbox App

There are probably entire engineering teams for each micro-service in the overall product (personalization, playlists, player, etc.).

That's a lot of apps to maintain and keep up to date with the latest features.

I would presume it's many more products than just Spotify app.

No Spotify, I do not want to listen to Joe Rogan and other right wing podcasts.

Just let me hide Podcasts already :-(

Spotify was a stand-out, almost no large tech company bought into cloud like they did, and everyone said the same thing: "It's not our core competence, it would require more people".

I get it, it's not sexy at all to deal in infrastructure, but I've seen their cloud bill and it's significantly higher than 1,600 peoples jobs, even with the discounts they got through committed use and even when considering the actual costs for infrastructure they need.

I'm sure there are inefficiencies everywhere, but this was the one that I talked about before and was talked down in a rather condescending tone. "Nobody got fired for buying cloud" is evidently a misnomer, because if you spend a lot of money and you don't have a lot of income: something has to give. And cloud has a lot more lock-in than most employees.

edit: And that moment depends not just on costs but the growth rate of the company so it's almost impossible to predict ahead of time (otherwise wall street investing would be a lot more boring).

> Spotify is cutting almost 1,600 jobs as the music streaming service blamed a slowing economy and higher borrowing costs in the latest round of redundancies at big tech companies.

It clearly states later that in a zero interest environment Spotify borrowed heavily and over hired, in this environment it’s no longer sustainable and they had to let employees go.

Did you just assume “costs” meant infrastructure costs like cloud?

it's easier to lay people off than to save millions in cloud spending when you've already committed to $x-billion over 5 years.

You extrapolated that costs meant cloud costs vs. them overhiring during the pandemic and investing in things like podcasts, which haven’t had the expected returns. You did this in an attempt to tie this back to some injustice that happened to you in a comment section far far away, which you’ve edited the post to remove now, thankfully.

1. I didn't edit my comment to remove anything like that, so I'm not sure what you're suggesting.

2. I didn't perceive injustice, I thought it was a bit brainless to not associate company costs with long term survivability.

3. An engineer being 400k TC is an anomaly, Spotify does not pay any of it's Swedish engineering force nearly that much, and since we don't know the demographics of the layoffs it's hard to argue engineers anyway. (Citation here says the avg was 125.000[0] which is still very high if these were europeans)

4. Their cloud bill would still be roughly 500 people based on that TC based on committed use alone.

5. Costs = Costs. Overspending in many areas = no more money. I'm merely suggesting that they overspent in one area that is now affecting another area indirectly. Yes, I drew a line.

[0]: https://news.ycombinator.com/item?id=38516794

Are we going to ignore that on average their US engineers costs significantly more than ~400K total costs and they have an engineering hub in New York City?

> 4. Their cloud bill would still be roughly 500 people based on that TC based on committed use alone.

And they paid Meghan and Harry 20,000,000 USD for a podcast deal that didn't work out. That's 50 engineers. Should we look at the Rogan deal too and go through everything that didn't work out in the last 3 years?

You've provided no evidence that by not using the cloud Spotify would have made more money or prevented these layoffs. And costs is one thing: Could they have grown as quickly? Scaled as fast? Hired as easily? Without numbers, which you won't share (outside of "trust me I've seen it"), this becomes even harder to debate.

Yes, because you're asserting that:

A) That engineers who have been laid off have that TC.

B) That it's engineers we're talking about.

C) That it's mostly localised in NYC; a high CoL city by all metrics.

Their other spending is also shameful, it doesn't discredit other poor spending.

You're a little bit upset with me for some reason, I would surmise that you're feeling somewhat defensive, maybe you work for a cloud or you've skilled entirely into only being able to work with cloud. That's fine, but you need to understand financial constraints in business.

This is the side of business I am most knowledgeable about so I am qualified to have an opinion; situations like Harry and Meghan? Rogan? Not my area at all and it would be impossibly arrogant of me to assume I know anything at all about those situations.

Also: "trust me bro" is not my position, the way I saw the numbers is a grey area legally and I'm not sure any Spotify people want to chime in to clear it up because it's likely one thing that is under tight NDA.

Here's some very old info though that'll help you realise that the order of magnitude is there: https://www.cnbc.com/2018/03/20/spotify-will-spend-nearly-45...

My comment you are referring to.

For the record, I used Glass Door, checked mostly Stockholm, but Paris, Barcelona, and Helsinki all seems to be in a similar range.

which lists most roles (including managers and senior engineers) as making less than SEK 1M ($98k) per year, most in the $75-$87k range it seems. Add on office space and equipment and other costs of roughly $30k per year and you got something like the above number.

$500/mo for seat licensing (for developers) this includes: Git hosting/Copilot/Office Suit/IDE's/Figma/Slack etc;

$500-$1500/mo for office space (depends a lot, in my company a seat costs me $500/mo and I'm in the middle of the city- larger companies may pay even less but lets assume a very high cost since it's a tech company).

$50/m for preventative healthcare (friskvard in sweden; this is actually higher than the maximum tax-free value)

$100/m for non-preventative health insurance (IE; private healthcare)

$120~/m amortised cost of laptops (assuming $4000 macbook over 3 years and a screen amortised over 5)

5% monthly salary for pension, so, that varies a lot.

But it roughly aligns with your $30k p.a

They could’ve been more careful about avoiding lock-in by building more on something like kube but most startups don’t have that foresight, and the expense of moving to on-prem is compounded even more when they have so much wrapped up in the aws ecosystem.

You won't get any argument from me on that one, but it's worth remembering that it's also hard to find people who are able to run systems in the cloud cost effectively.

I believe this means they generated $500m for bytedance, but I can't find a good citation, bytedance itself generated $6B in profits though for the fiscal year 2022

https://www.statista.com/statistics/1342785/bytedance-key-qu...

No, but you're definitely dealing with the Arthur C. Clarke quote: “Any sufficiently advanced technology is indistinguishable from magic”.

When I used to work in a similar environment we would develop code, then we would give it to a QA team. They would test it and give it to an Ops team. The Ops team would schedule a maintenance window and roll out the new code on each server. This happened maybe once a quarter because testing and releasing was a week-long process.

Racking new servers and provisioning them also required some manual labour. We had a process to use PXE to provision the machines but it was still toil. Virtualization was a big benefit because you could at least create and blow away VMs without having to re-image a whole server from scratch.

"Running your own cloud" implies that developers can treat instances like cattle and interact via an API. But it also means there's a standard set of tooling for fleet management. None of this stuff is entirely new but in small or mid-sized orgs it was out of reach 10 years ago.

Cloud: log onto a web portal, select the specs I want for the servers, press OK and log in 30 seconds later.

Oh wait, that's just the IBM Cloud...

Citation needed

>ValueAct’s chief executive, Mason Morfit, said Spotify’s costs had “exploded” and that it was “built for the bubble”

Isn't this what made Spotify success? Yes, they could have build their own datacenters, but someone else could outcompete them in the meantime.

They have traded off something for another thing.

Absolutely, but everything has ROI, ROI of fixing cloud bills were probably smaller than "possible growth"

Now they know which features failed, they can start deprecating or stop growing them and refocus on cloud bills

The question becomes then: 9,500 employee's and none among them know how (or wanted) to regain enough costs to save 17% of the company from losing their job just before Christmas?

Certain features (spotify wrapped for example) would definitely be a lot harder with more traditional hosting. But the bulk of Spotify's workload are not significantly hampered by moving slower on infrastructure and recuperating significant costs.

It's just very easy to keep piling infra costs on top of infra costs and not thinking about it, especially if it feels disconnected from company financials for most developers and they do not see a line between spending and jobs.

> I've seen their cloud bill and it's significantly higher than 1,600 peoples jobs

An apples-to-apples comparison requires looking at all associated costs of switching to on prem (salaries, hardware, etc...).

I've only been a part of this analysis at a company dealing with a cloud bill in the low tens of millions, nowhere near the scale of Spotify (and I'm still relatively new to all of this).

> And cloud has a lot more lock-in than most employees.

It's a shame there are not better abstractions to facilitate moving to and fro.

For ML workloads anything between 6-8x cost increase for using cloud.

For CDN costs somewhere like 3x (they're pretty smart on discounts here, without those it could be 100's of x)

For compute somewhere like 3-4x (would be 6x before the discounts)

Managed DB is a hard one because the cost is really high but I'm not 100% sure how much time is saved with the solution they chose because while I saw the bill I don't work there.

It gets dicey with humans because it's apples to oranges, expensive SRE's who understand the inner-workings of cloud vs cheap(er) hardware folks who set things up from first principles with a higher up-front cost. -- FinOps with more traditional solutions is easier but you still need people for that.

Spotify's backend load is not entirely elastic. There are some elastic bits but a large chunk is pretty static.

I'm not sure what you mean by this as Spotify was actually quite late buying into cloud. They ran on bare metal through at least 2015. From their engineering blog:

">Thus, in early 2015 we started exploring what a cloud strategy would look like for Spotify."[1]

So they would have been running on bare metal through some of the biggest spikes in growth i.e launching in the US market. I don't think they bought in any more or any less than anyone else who transitioned from bare metal to cloud and for probably the same reasons.

[1] https://engineering.atspotify.com/2019/12/views-from-the-clo...

> “Embracing this leaner structure will also allow us to invest our profits more strategically back into the business,”

> invest our profits more strategically back into the business

why didn't they do this in 2020 when they got zero interest loans and free money from the government?

profits are for "strategic" investment but loans are for un-strategic and unsustainable hiring? got it.

>> This is not a step back; it’s a strategic reorientation. We’re still committed to investing and making bold bets, but now, with a more focused approach, ensuring Spotify’s continued profitability and ability to innovate.

"Strategic" could be replaced with "efficient". They were previously optimized for growth -- now, they're optimizing for growth:krona.

Also, it is refreshing to see a head cut announcement that doesn't bury the lede: we're firing people, and this sucks.

"We optimized for one thing. Things changed. No we're optimizing for a different thing."

Doesn't need to be more obfuscated than that.

In lieu of actual profitability, they also chucked in $1bn to buy up their own stock to maximize the CEO's already enormous compensation.

https://techcrunch.com/2021/08/20/spotify-to-spend-1b-buying...

It's so hard to fire people in France, that some large companies simply have a policy of "no france office".

If a business cannot afford to do business in a country, then that business should not be allowed to do business in that country.

https://data.worldbank.org/indicator/NY.GDP.PCAP.KD?location...

US GDP per person is now 50% higher than france:

https://www.worldometers.info/gdp/gdp-per-capita/

It wasn't like that 15 years ago.

There doesn't need to be another rule. The existing rules are doing just fine keeping companies out.

Layoffs are not desirable for a company, but sometimes they are required. The fact that a CEO decides to go with layoffs when it's the right thing to do shouldn't be seen as a failure of the CEO. It's the pit stop that keeps the company from exploding if you just ignore all financial health markers and keep everyone employed until the money runs dry and you have to gut the business and sell of part to cover the bankruptcy.

Laying off 1/5 of staff is not a "pit stop" to refuel. It's an admission of staggering mismanagement.