Hacker News Reader

Comment by fixermark | Hacker News Reader

Worth noting: the article's headline is one of those tricky situations where the summary isn't wrong, but should probably include more information.

FB allows advertisers to target at specific topics, and they've been blacklisting objectionable categories. But the blacklisting appears to be manual, so while "nazi" isn't a micro-targeting category, things like Josef Mengele and a white supremacist punk band are.

Manually keeping up with and out-thinking objectionable content keywords is a perpetual arms race. If FB wants to win it that way, they'll have to invest pretty hard in that space if they don't want a story like this every quarter.

View on HN

Why not just ask people for $1 per year to use FB, and then have non-targeted ads to cover the rest of the loss in revenue?

Because an insanely high percentage of people will refuse to pay $1/year. Also, because $1 does not come close to covering the cost difference; targeted ads typically bring in 2x or more revenue per impression and the average revenue per user per year in the US is around $25.

I don't know about that. The latest estimates I heard was that you would have to pay the average American $1k not to use facebook for a year.

That doesn't mean they would or can spend $1000 on FB, but with the proliferation of subscriptions on app stores that are $10+ a month, I bet FB could get $10 a month from a large chunk of users and start to move towards other revenue models.

Ah but that is while it is free and presumably everyone else is using it. If it costs a dollar people would have to enter payment information all of a sudden, so it is putting a decision point in front of people where they have to expend effort and money (albeit a small amount of both) to keep using this thing.

If their friends seem hesitant about it maybe they won't sign up to pay for it.

This isn't to say they could make as much money with such a model, but if they were more concerned with the downsides of surveillance ad targeting capitalism as their revenue stream, they could certainly explore shifting to be a different sort of company - one with different and maybe even less revenue streams.

Funny thing that, why do we never hear about a company that wants to stay the same size or shrink? What is facebook decided they wanted to be half the size they are now expense-wise, get out of the ads business, and encourage people to use their platform for less things?

> Funny thing that, why do we never hear about a company that wants to stay the same size or shrink?

I guess it's a rhetorical question, but if you went public it would scare away all the investors, right? Who would invest in a company that does not plan to grow, and might even shrink?

Private companies sometimes do that, tho.

targeted ads typically bring in 2x or more revenue per impression

Source?

https://www.researchgate.net/publication/265266107_The_Value...

Study quotes 2.68x earnings in 2009 but in my experience it’s even higher these days.

Does that mean advertisers pay more for them or that they are actually more effective (or both, or other?)

Their operating costs would be massively lower if they didn't have to collect every inch of mouse movement etc. for targeting advertisements. So even with lower income, they might still be as profitable or more profitable.

Facebook made $48B in revenue last year. That's around $38/user (with much more coming from people in North America). Even $1/mo doesn't come close to covering it.

$1 is vastly different from free for a variety of psychological reasons. People will also be less likely to want to pay to see ads on every other post. You'd have to provide more value. Also, the fees from credit cards companies are so high that micropayments don't make sense (yearly subscription charge + per-transaction fixed fee + per-transaction percentage charge).

Because that wouldn't cover the loss in revenue. $1 per year and non-targeted ads doesn't equate to $50B+ in revenue growing 20-40% per year. Also, advertisers come to FB because they can be ROI positive with the targeted ads. Most advertisers don't want to run non-targeted ads because they would be ROI negative.

Zuckerberg would become Steve Case wealthy.

Facebook at $X/mo is AOL. AOL was a successful but fairly predictable business. But Facebook offers a promise of growth that promises more return.

most ppl in the world have no credit card. I didn't have one till i visited the west.

It's a 'bad headline' if it will be commonly understood as something not representative of the content.

It's done on purpose by all publications, because the nuanced reality of most situations isn't inflammatory enough to drive clicks and create commotion.

Given this situation i.e. wherein it's really hard to track down and manage the colloquial lingo used around the world for various things ... the headline is unfair.

"Facebook fails to stop advertisers from targeting some extremist memes"

"Facebook unable to tamp down extremists shifting lingo"

"Nazis by another name: advertisers target extremists using shifting terminology on Facebook"

There is definitely a more responsible headline here, and surely the editorial staff are capable of that if they wanted to.

Surely the staff at many publications are wary of this as well, it's one of the ugly pressures of business reality that 'someone' is enforcing.

As for FB ... this has to be hard, whack-a-mole kind of things. Sometimes I'm sympathetic to them, other times I think $100B and some of the best AI folks in the world should be able to mostly figure this out.

Much like rampant fake goods on Amazon ... Bezos can land a rocket by itself, but can't get counterfeit goods off of Amazon? ...

Not that I'd defend either company, but to be fair, landing a rocket is applied engineering and math. Invest enough to build it right, and it will land. Making subjective calls about "good vs bad" content or products, where what's considered good or bad may ebb and flow with current political/cultural whims, is not so straightforward. While "are nazis bad?" is pretty black-and-white, there are a lot of gray area topics. Not sure if adding magical AI fairies would help.

I disagree with both points.

"Counterfeit goods" are 'black and white'. Not 'maybe good or bad'. They're either counterfeit or they are not. Maybe tricky to track down in some cases, but ultimately - objective.

Yes, stopping it might probably be fairly manual process involving process and overhead, but it's achievable.

Bezos spends $1B a year on his rockets - you don't think he can tamp down counterfeit goods for that amount?

I'm always shocked at people defending Bezos flagrant hustling of counterfeit goods. He can stop most of it, he knows about it, therefore he implicitly choses not to.

...

'Nazis = bad'? That's easy.

But this is not the subject at hand. You're missing the issue that the headline is misrepresenting the subject! Facebook has banned ads for Nazis, because that's easy.

We're talking about ads for 'Skinhead Bands' - or other ill defined groups. Well, Skinhead is a complicated subgenre, it's not black and white. But ok, ban that in ads. What about 'Reggae Skinheads'? 'Black Metal' - wait, is that a Nazi thing in Europe, or just basically 'metal'?

And what about terms in France? Ukraine? Algeria? Every culture has a broad set of terminology, sub groups and political dynamics that make this thing fairly difficult to interpret.

More important is the underlying fact that it's subjective. Nobody can really satisfy all voices because we all have different notions of what 'wrong/evil/bad/offensive' is.

Just like there is effectively no way to ban 'hate speech' on FB because there's no globally accepted definition of what that is, kind of the same applies to ads.

They can stop Nazis, but maybe not some weird 'bad' skinhead subgenres that popup now and again.

Counterfeit goods are a good example where it's not black and white. Say Louis Vuitton runs a factory, and it's workers sneak back into the factory, make a couple of handbags and then sell them themselves, literally the exact same product as Louis Vuitton sells, but isn't sold by Louis Vuitton, I would say that is still a genuine "Louis Vuitton handbag", on the second level, whether or not counterfeiting is bad is also not black and white, counterfeiting makes it harder for companies to sell their products, but also reduce prices for consumers by increasing competition.

Reply on news.ycombinator.com