60 comments

[ 4.3 ms ] story [ 120 ms ] thread
It's wild that Youtube makes $8.7B in ad revenue a quarter, and adding in another $15B from subscriptions it is touching on $50 billion in annual revenue.

Which is more than:

* Netflix – $36.3B

* Warner Bros Discovery – $40.57B

* Paramount Global (ViacomCBS) – $30B

* Disney Entertainment (excluding theme parks and sports) – $40.6B

All without spending a single dollar on content creation.

YouTube advertising revenue: $8.66 billion vs. $8.93 billion expectations

Google Cloud revenue: $10.35 billion vs. $10.20 billion expectations

Percent increase in YouTube advertising revenue from Q2 2023 to Q2 2024 approximately 13.02%.

Percent increase in Google Cloud revenue from Q2 2023 to Q2 2024 approximately 28.84%

Edit: Updated to reflect correct numbers...

It did not decrease. $8.93B was what analysts were expecting.
Thanks for pointing that out. I will update my post.
> All without spending a single dollar on content creation.

YouTube paid out $70 billion to "creators, artists, and media companies" over the past three years.

https://blog.youtube/inside-youtube/2024-letter-from-neal/

I don't think people are aware that the majority of a yt premium subscription goes to creators. Premium views also pay more than ad views to creators.

Not that I want to shill yt premium, but youtube overall would be dramatically better if google was beholden to users rather than advertisers.

> I don't think people are aware that the majority of a yt premium subscription goes to creators

Huh? Maybe Google can claim this through some very creative accounting, but the bottom line is that Google retains the vast majority of YouTube revenue; and that the vast majority of content creators on YouTube are not getting paid at all, because the requirements for monetization are pretty steep.

That last part is deliberate, by the way: it allows them to keep a larger slice of the pie while running ads on all videos and showering a small sliver of celebrity content creators with piles cash. For me as a content consumer, YouTube is fantastic. For the average content creator, it's a miserable experience largely built on a lie.

> but the bottom line is that Google retains the vast majority of YouTube revenue

Do you have a source for that "bottom line"?

In this thread, the numbers cited (I didn't verify) are: - $50B/year revenue - $70B/3 years in payments to creators

Youtube paying out ~50% of revenue to creators seems almost too good to be true to me. That isn't their only cost, the infrastructure of youtube must be insanely expensive. They're almost certainly taking far less in profit than they are paying to creators.

edit: I read the numbers in the thread wrong, $15B/year in subscriptions not $15B/quarter.

Ah that seems more reasonable, with $8.7B a quarter in advertising revenue, that's about $35B a year, so for Youtube annually:

  - $35B in ad revenue
  - $15B in subscriptions
  - (minus) $23B paid out to creators ($70B / 3 years)
Which seems to line up with around 50% paid out to creators ($23-24B out of $50B revenue).

I don't know how many employees work on Youtube specifically, but even with salaries and infra / CapEx / OpEx costs, they are probably making a decent profit on the remaining $25B they keep annually after payouts.

The most scammy and evil move from them was monetization of content from non-partenered channels[0].

[0] https://www.cnbc.com/2020/11/19/youtube-will-put-ads-on-non-...

I don't get it. They host the videos for you, why shouldn't they recoup their hosting/traffic costs somehow?

Scammy? Evil? Damn...

I didn't agree to have ads from X or Y company roll before my video/s. If they don't want freeloaders then make it subscribe only service. They leech data from us freeloaders why wouldn't we leech hosting resources from them?
> I didn't agree to have ads from X or Y company roll before my video/s

Maybe read terms and condition before loading videos on youtube?

Guess what? They change their TOS constantly meaning; I agreed to the initial TOS but I didn't agree to the subsequent ones. You can argue, "Yes you did, you clicked that agree button and you kept using it". Yes I did but only because they have the biggest video community on the internet, which means the biggest content offering and the biggest reach. If there was decent competitor, I would move immediately.
Spoiler: A competitor would be just as shitty to you. No one wants to cater to people who think showing ads to support the costs is "evil and scammy".

Actually, vid.me did want to cater to people to you. And guess what happened? They went bankrupt once their VC cash was all burned up.

Read my op; I said showing ads on non-partnered channels is evil and scammy not ads overall. As a matter of fact I watch ads on my YouTube mobile native Android app.

All I want to say is; if I have a small channel and if I am a small/hobbyist content creator, please don't show ads on my content because I didn't explicitly agree to it.

Well you can go ahead an make a video host that provides free bandwidth and service to people who "don't explicitly agree to show ads".

It will be extremely popular, and you will become very...broke.

That's the creators' share of ad revenue, and Google pays it out only after it has earned the money. If the video doesn't get views, the creators get nothing. It's the same as Google putting ads on your website. You get a share of the revenue, but that doesn't mean Google paid for the site.
The point still stands that YouTube does pay for content, it just does so in a semi unique way.

I say semi unique because revenue share is common in traditional content creation as well. So it’s not like YouTube is completely alone here , though it is rare for someone to go purely revenue share.

That said, YouTube has funded content creation directly as well in the past https://en.wikipedia.org/wiki/YouTube_Original_Channel_Initi...

I disagree. Paying for content, whether directly or indirectly, and being a platform that connects users/advertisers with content owners and takes a cut of all transactions are two very different business models. YouTube is very clearly the latter.
I think it's more accruate to say that Google has done all of this without investing in content
Those are pretty crazy stats! Definitely impressive, hope the parties responsible for making it happen get their fair share of the spoils.

It's also wild people still watch youtube so much. Ever since Elgoog dialed the YT ads up to 11, I'm always hesitant to watch anything without an ad-blocker (e.g. casted to a TV), and spend minimal time there these days. Previously I didn't bother with ad-blockers because the frequency was reasonable.

I feel like they violated the pact of not turning it into a cable-tv like experience. I definitely no longer keep up with any of my subscriptions.

Guess I'm not the target market :) No surprise there, hehe.

I have yt premium/Red, whatever they call it, and it's worth it, I think. It's always jarring when someone else uses youtube and I'm flooded by commercials.
I have yt premium because I want to pay the creators I watch. And yes logged out non YTP is view into another world.
I have it to keep my kids safe. The family plan has gotten expensive but still worth it to ensure they dont get those revolting adverts.
I watch a lot of YouTube during the winter. I use uBlock Origin on my PC and never see an ad. There's a ton of useful content related to my hobbies.
I combine uBlock with NoScript, and it’s life changing. YouTube has never loaded faster lol
25% of youtube views are music, so add the revenue of Spotify to these, and it adds up
“Only a moron would buy YouTube”, Mark Cuban, 2006
He made his fortune on broadcast.com, notable only for its acquisition price.
It's an odd feeling to have seen the rapid decline of QoL on the interior, know it was to drive margins back above 30% / $25B/quarter, leave because of it, then profit from it. Feels unethical.

I guess that feeling is just having the dissonance of the nature of business in your face. Though I do hold that the decisions made to squeak out a couple extra % points will be crushing in the long run.

Even just today: reality is Google is 4th place to OpenAI, Anthropic, and Meta in AI and its not really close.

Each individual step is arguable, but the cracks are there. And leadership is too MBA-y/trend-following to do anything about it. Ex. through August '22 the line was there was going to be a special pot of increased raises to combat all the talent leaving for Meta. Then once Mark started firing, all that went far away, promo quotas to manager level were dropped to ~0, there's ceremonial cullings weekly, and its not of managers.

Edit for child comment:

I respect what you're saying: I'd advise its better to invite conversation, than say its all a simple binary option, especially in the face of me pointing out the same thing.

Ex. morale is trash, outsourcing is rampant, virtually all new headcount is overseas, and _actual_ efficiency, productivity, is lower than it ever has been.

The more simplistic view would advise companies could just fire 90% of staff and go into maintenance mode, but for some reason, they never do.

Also, note that "efficiency" on HN arguments of this form means "get the managers / slackers out of the way" -- economic effiency means "profits are 0". Neither happened here. Never been a better time to be a manager, and slackers aren't going away, if their slacking was visible to managers, they would have been gone already.

It's called supply/demand and living in reality.

If GOOG stocks drops 30% relative to it's peers, a majority of them will leave ship.

So, you can't reap the benefits of increased stock price and yet complain when companies get efficient

The complaint is not that Google is too efficient. It is that Google is achieving apparent improvements in efficiency by taking steps that improve the short-term metrics and have net negative long-term impact. So, OP is disputing the senior management claim that this cost base reengineering is legitimately "durable."
The cost savings have been bad for Google. Reduced productivity significantly. A lot of things are failing behind the scenes. A lot of the revenue lift is coming from short term hacks that will not lead to sustained growth.
SREs and other Engineers have an mismatched sense of importance of what really matters for running a profitable business.

And given a choice they will always invest and advocate more hiring, more quality, more gatekeeping, more process

You're applying a Current Thing (2 years old and still kicking!) template on something that doesn't match the template.

This will be the 3rd time you're hearing from a Googler that this made gatekeeping and process significantly worse. It's not straightforward to imagine why, and given your interest in engaging in conversation, it's certainly odd to not be curious why.

Once you understand more, you will appreciate this template you enjoy even more. No reason to cling to in every situation, repeating it in a boring way that makes you look boneheaded.

from my interactions with Googlers, they are pretty bad at assessing non-leetcode related strategy.

E.g Ask 10 Googler Engineers and at least 80% think AI is hype and left to them they will underinvest in it.

Another random comment. I left Google to do an AI startup. I agree with your general observation and have many highly specific hair-raising observations to back it up. I cannot tell based on your erratic contributions if you are GPT 3.5.
Hopefully the death of Google will provide sustenance for a new wave of companies the way a whale fall provides for the scavengers of the deep.
If you think one of history's largest companies, inexorably woven into the fabric of the digital age, is going to "die" in any meaningful way soon, you need to reevaluate your model of the world.

I have lots of criticisms about the people in charge of Google products over the last decade, but it's a very long way from the stratosphere to the ground.

What is your experience with how Google operates internally?

The finally seen from Thelma and Loise, that is how Google ends.

(comment deleted)
I just listened to their earnings webcast. It looks like it's going to be yet another earnings release season of tech companies spending billions on AI infrastructure with nothing to show for it. And they're promising continued levels of intense spending on compute. Sundar's comment was that it's better to overspend than underspend and fall behind the competition. Time to buy more NVIDIA.
The time to buy Nvidia was 6 months ago. At this point investors have already priced in many years of exponential growth.
The time to buy Nvidia was 5 years ago, and you would have had 30x gains (or even earlier). But the upside now does look more limited
> with nothing to show for it

AI and LLM features are everywhere in Google products, on top of just using Gemini itself.

I have no idea what you mean. They have tons already to show for it, and presumably this is just the start.

Their revenue grew 14% YoY. That's nothing special at all if you look at their historical top line growth.
Google (Alphabet) is already the 4th largest company in the world by market cap.

So, what -- you want to see their revenue grow by 50% or something in a single year?

For mature companies of that size, that's not really a thing. I think your expectations for defining AI success may be a bit too high here.

Are you kidding me? 50% is peanuts given that AI has been touted for years to be more profound to humanity than electricity or fire:

https://www.cnbc.com/2018/02/01/google-ceo-sundar-pichai-ai-...

AI is supposed to be a huge new revenue driver and my point, echoing others, is that it's nowhere close. They've spent double on AI compared to last year:

https://finance.yahoo.com/news/alphabet-earnings-top-estimat...

"In the second quarter, Alphabet reported spending $2.2 billion building AI models across its DeepMind and Google Research organizations. That's up from $1.1 billion in Q2 2023. When exactly AI starts to generate revenue for Google’s Cloud business, let alone its ad segment, is still up in the air."

Interesting that investors don't care at all about the Waymo progress, judging by the q&a. Though, Tesla is definitely an AI market leader!
That's cause so many of them live in NYC and are oblivious to things that don't exist there
Promises from a serial breaker of promises somehow justify triple the multiple of actually delivering a real product that does exactly what it says it does.