Hacker News Reader

Comment by spondyl | Hacker News Reader

I always wondered why tech companies had such an impenetrable inward communication until I came across the following insight:

"If you have a billion users, and 0.1% of them have an issue that requires support on a given day (an average of one support issue per person every three years), and each issue takes 10 minutes on average for a human to personally resolve, then you'd spend 19 person-years handling support issues every day.

If each support person works an eight-hour shift each day then you'd need 20,833 support people on permanent staff just to keep up. Amazing."[0]

It's not necessarily the exact same source that I discovered back then but the idea is the same. Even a tiny portion of customer service would still soak up a large amount of money, perhaps rightfully so, but at the scale some of these companies are, it quickly becomes unsustainable.

[0]: https://plus.google.com/+DeWittClinton/posts/1hRWj489oEz

View on HN

It's a misleading estimate because it doesn't define "user". In the case of YouTube, there are a billion users but only creators would require support.

Maybe there are 100,000 YouTube creators who get paid anything meaningful? (I have no idea, just guessing.) If 0.1% of them require support, that's 100 requests per day. Any moderately successful mom-and-pop SaaS gets that much.

There's no financial reason for Alphabet not to provide personal support to YouTube creators. They just don't want to.

What you're suggesting is how it already works. Partnered creators all have access to a creator support email (with a 1 business day SLO). Top partnered creators (aka the "creators who get paid anything meaningful") will also have a partner manager they can contact for 1-1 advice/support/etc.

Great. So maybe the bar for support is set too high? Someone making $50 / month on YouTube is meaningless to Google but may spend an enormous amount of time on that content in the hope of a breakout. Those are probably the people whose mental health is most at risk.

It’s Alphabet’s community, they own it. Extending more support to dedicated members may not make direct business sense but it could be the right thing for long-term community health.

> ... for long-term community health

I understand words like books and bricks and apples, because these words have physical referents. I have no idea what "long-term community health" means.

What does that mean?

And why would anyone invest more in it than in barricades and highly weaponized security guards? (I'm not actually promoting an armed response to whatever a 'community' is, I'm just using it as a rhetorical foil to try to understand the pros and cons of investing in whatever a 'community' is)

A community is a group of people you interact with (and who interact with each other) regularly. Community health means the people are kind to each other, help each other, communicate with each other. If a community is not healthy, they don't talk to each other and view each other as enemies.

If you have barricades and highly weaponized armed guards, you don't have a community. You have a war.

> If you have barricades and highly weaponized armed guards, you don't have a community. You have a war.

Or a highly pacified citizenry? ("Move on citizen. It's after dark. go home.")

What role does Youtube have in making a kumbaya kind of experience?

>Or a highly pacified citizenry?

That's still a war, just a very one-sided war.

>What role does Youtube have in making a kumbaya kind of experience?

I'm assuming the question is "what reason does Youtube have", since their role is pretty clear: they control the platform. If they want people to keep using it, they have to keep those people happy. If they want to prevent a revolt, they have to keep people happy. Because unhappy, unstable people can turn violent. Keeping them happy won't necessarily stop all unstable people from turning violent, but you certainly do maximize that risk when you have some unstable people in your community and your community is already on the brink of a user revolt.

What's the percentage? If 80% o the peasants are revolting, and 80% of the content comes from the unrevolting peasants, do I care?

I'm not advocating that they should or should not care, but if one percent are revolting and 99% comes from the other, don't you think it is time to bring out the uzis?

> What role does Youtube have in making a kumbaya kind of experience?

Why would any corporation want its customers to be happy?

Some businesses certainly don't need to care. A restaurant in central Rome has an infinite supply of new tourists to feed, so they don't need to worry about whether anyone wants to come back. At the other end of the spectrum, a monopolist knows that customers have no choice and must come back regardless of how they're treated.

YouTube has elements of both: there's a near-infinite supply of hopeful Internet video stars, and they also have a near-monopoly on video viewership. They might still want to keep stakeholders happy because these issues tend to brew for a long time until they blow up into public consciousness. (Facebook's promiscuous spreading of user data was known ten years ago, but they only took action now. Damage control would have been easier earlier.)

That is a very short sided perspective. People will ask local tour guides for recommendations, and if they say you suck, you have lost business.

But how do you tell those dedicated users apart from all the bot accounts that post millions of hours of garbage backed by a computer voice narrator?

Bot accounts won’t be calling customer support.

Creators have these questions - "why is this seemingly harmless video demonetized?", "why are my videos not reaching the people who explicitly subscribed to me?" and support won't answer those. They blame the algos and point to you a vague list of guidelines.

>why are my videos not reaching the people who explicitly subscribed to me?

Ugh Facebook. I have a community page for my town with thousands of likes, and most posts only reach 10% of the people who explicitly followed my page. Even if I pay to boost a post, I'm lucky to hit 20%. I get a lot of complaints from people who missed an event because they never saw the (sometimes multiple) Facebook posts announcing it. I mean, they could have visited the website or subscribed to the newsletter, but still... they asked Facebook to kindly show them the posts from this page, and Facebook told them "no", because "algorithm".

computer says no.

> Partnered creators all have access to a creator support email (with a 1 business day SLO)

Does that still apply if they get summarily "de-partnered"?

If a company can't support the number of customers/clients it eagerly takes in, it's time to either scale up support or scale down customers/clients.

Which they are doing in a way, by demonetising them.

Plus by screwing their current clients, they are in effect working towards not getting new ones. It's a win/win scenario I suppose.

It's not "screwing" them, you said it yourself. They can't provide adequate support to all of their customers (in this situation, customer is defined as a content creator), so they are reducing the number of customers that they take on (by making it so you need a specific number of views-per-month or subscribers before you are eligible for monetization), and are reducing the number of current customers they have (by demonetizing the videos and categories that advertisers like the least)

Great solution give or take the occasional shooting. The shootings don’t cost much and most shooters can’t scale their killing so I guess it’s all good.

I don't know why I'm even bothering to respond to this, but what is your proposed solution that keeps the company in-business and ensures no mentally-ill people decide to shoot up the company?

1. Hire more support

2. Provide clearly defined guidelines as they relate to the algo and be equitable with enforcement

3. Be politically neutral

This will not guarantee safety from violence, but it could increase it.

1. Doing this could basically cause the same result, as the cost of more support could easily make it so that they need to start charging or denying who can be a content creator. Neither of us have numbers, so it's all just guesses, but I'd be willing to wager that spending a TON of money and training on new CSRs wouldn't really change all that much, especially when telling someone explicitly "you were banned because X" is almost never a good idea, all this would do (in my opinion) is create a much more expensive, slower, and more annoying version of the same problem.

2. I think this would help, but it's not going to prevent mentally ill people from being mentally ill. They aren't going to say "oh well this removal was justified and was consistent with the others", they are going to find a perceived wrongdoing and will latch on to that, because they are mentally ill.

3. I've come to the conclusion that this is literally impossible. You can't be politically neutral. People aren't politically neutral, and therefore the things they create or moderate can't be politically neutral (whether they mean to do it or otherwise). Even algorithms that are created by people can show biases.

>You can't be politically neutral

No, but it's pretty easy not to be censorship happy, far left SJW activists.

WRT #3, the problem seems to be moreso consciously executed covert and overt political censorship/demonetization of undesireable speech rather than the problem of unconscious biases of developers. It's as simple as refraining from censorship

But when "refraining from censorship" ends with advertisers (your main source of income) pulling out, your options become "censor videos" or "shutdown the service".

"Not playing the game" isn't ever an option here.

>But when "refraining from censorship" ends with advertisers (your main source of income) pulling out, your options become "censor videos" or "shutdown the service".

False dichotomy, there are other business models that don't require selling out to the whims of advertisers.

And YouTube is trying out those other business models, however the extreme vast majority of users don't want to pay, and the content creators already have the option to self-host, self-fund, or even use YouTube un-monetized and setup and run their own in-video ads.

YouTube is mainly an ad supported product, and just because other business models exist doesn't mean that you can call curation of the content "censorship".

This would be the exact same situation if say a website used a "subscription" model where you paid per month, then the credit-card processor decided that the content was "unethical" and dropped the video service as a customer. And that's not theoretical, it's happened to many porn hosting websites out there.

"Not playing the game" isn't an option, no matter what at the end of the day you need to "censor" some stuff (even if just to avoid getting in legal trouble), and that "censorship" will always be more strict than absolutely necessary, as the risk for a "not removed but bad video" far outweighs the downsides of a "removed but ultimately not bad video".

We are in the golden age of information in my opinion. It has never been easier to self-host and self-fund your own content in a way that is basically uncensorable. Use it! But don't go trying to change or destroy other platforms that explicitly don't want that content, or want to pursue another business model. This isn't a zero-sum game.

2 is the key. It’s not just “demonetization” it’s seemingly arbitrary demonetization. It seems like political censorship to some people, to others it seems random and capricious. Nobody seems to have any clue as to how it all works and the natural result is that people take it personally. Transparency costs nothing and scales infinitely.

Well, I fall on the side of calling a situation involving a company promoting a service that is monetized and then suddenly (sometimes quietly) removing said monetization from people with little or no attempt at justification with strange/vague rules that no one understands nor will the company bother to explain as screwing your clients or customers or creators or whatever we wish to call them.

But I admit that could just be me.

But Google is explaining their actions here, at least on a macro level.

The blog post at [0] as well as an email summarizing the situation was sent to every channel under the new threshold for monetization.

And any time a video is demonetized, they are given links to various help articles like [1] that go over what isn't allowed. You are also sent a link on how to appeal the decision at [2] They don't speak to an individual reason or reasons, and likely never will, because it's too easy to game that. You end up with people following the "letter of the law" but not the spirit, then if you decide to ban them anyway, you get called out for applying rules inconsistently.

There's a reason why our legal system has judges, lawyers, juries, and takes many years in some cases to come to a conclusion. This shit is extremely difficult to manage at any scale.

[0] https://youtube-creators.googleblog.com/2018/01/additional-c...

[1] https://support.google.com/youtube/answer/6162278

[2] https://support.google.com/youtube/answer/7083671#appeal

What's so outlandish about that support organization? Presumably you have the revenue to match when you have a billion users. Walmart has 2.3 million employees.

It’s fascinating how the tech community is so obsessed with scaling and automation that the idea of having a five-digit support staff for a billion users is seen as absurd. Companies like Google have taken this attitude to heart and seem to be absolutely determined to avoid anything that might require a linear relationship between userbase and staff numbers.

Even if those statistics are accurate, my response to such companies would still be “boo hoo”. If your company wants to rake in the billions of dollars in benefits, you need to spend some money too. If your company can’t handle the time/effort then you shouldn’t aim for a billion users.

20,000 people doing support at a generous 10$ an hour is ~400 million a year. Which sounds bad but is only ~40cents per user.

However, 0.1% per day and 10 minutes per call is a very high estimate.

Also 1 billion users may be true for viewers, but the real support is only needed for content creators. That number is _significantly_ lower than 1 billion.

One good chatbot could probably solve more than half of those issues.

And drive the other half of people to insanity. There is, in support terms, not much worse than an obvious chatbot or support personnel that only has canned responses available. You explain your problem, they come with a canned response that suggests doing what you already did AND WROTE THEM, you reply back "I already, as you advised, did XYZ and wrote you, this does not help" and get the SAME canned response bullshit back.

For what the fuck did I just invest half an hour in a detailed bug report when you're not even bothering to literally read it?!

Google and Facebook have an employee:user ratio of roughly 1,000,000:1. This makes any hands-on intervention exceptionally expensive.

(Not impossibly, given high-enough levels of interest, but high.)

Google, and technology's, most powerful tool is its ability to amplify reach, capability, and efficiency. This works very well in a broadcast-style transmission.

Where that transmission is reversed, and the many nodes start trying to feed back to the single hub, there's a problem. It's interesting to note that much of the history of the formation of the modern corporate entity has been one of changing modes of communication (and the parallel development of management structures). Two authors who explore this in (well, for me) fascinating detail are JoAnne Yates and James R. Beniger

Yates: "The Memorandum as a Management Genre" http://www.ismlab.usf.edu/dcom/Ch6_YatesMemoMgtCommQtly1989....

Yates: Control Through Communication: The rise of system in American management https://www.worldcat.org/title/control-through-communication...

Beniger, The Control Revolution http://www.worldcat.org/title/control-revolution-technologic...

In particular, broadcast technologies give leverage, but work best, as with the agricultural model, where seed can be scattered at will, and harvesting is simple, bulk, and commodity. When your plants talk back to you and have concerns about their growing conditions and/or opportunities, you're going to need a more complex feedback system. And it will have to be effective.

This has long been an Achilles' heel of Google (and, to be fair, many other large corporations).

On which, see A.O. Hirschman's Exit, Voice, and Loyalty: Responses to decline in firms, organizations, and states. https://www.worldcat.org/title/exit-voice-and-loyalty-respon...

The fact that the modern corporation is, by its very intent and design, a risk-externalising machine, one that shifts danger to customers, users, vendors, and employees, rather than either the investor-principles or their manager-agents (and yes: principle-agent problem is another dynamic) is itself part of this very toxic mix.

> Even a tiny portion of customer service would still soak up a large amount of money, perhaps rightfully so, but at the scale some of these companies are, it quickly becomes unsustainable.

The problem is that now society has to shoulder the costs. For example, when a mentally not-so-stable person goes to YouTube HQ and starts shooting everyone in sight. Or, more common, people falling in depressions, losing their homes, ...

Reply on news.ycombinator.com