[untog]

I very much hope this isn't just words. The core problem, IMO, is that content that makes us angry, anxious or jealous is a much better driver of clicks than content that makes us happy. I'm sure Facebook knows this. If they really mean it, they'll accept that they will make less money as a result of this change. It would be the right decision in the long term, but the short term will hurt.

[jp555]

Are clicks the majority of revenue for FB? An ad unit that fills a users screen for a few second as they scroll throughout the day every day is super super valuable to brand advertising. I think that’s ~$450 billion of the total ~$500 billion total annual global ad spend.

To me browsing the fb feed is a lot like like flipping through tv channels used to be. Brand advertising loved that too.

I’d think the More people browse their feeds the more valuable their ad unit becomes to brand advertisers.

[Avshalom]

Right I mean, driving a click through to a website due to rage has to be less profitable than driving a click through to say DeWalt power tools due to delight. Right? I mean some fraction of a fraction of of an ad for DeWalt must be worth less than keeping your friends pro DeWalt status update on top.

(This is not a paid ad for DeWalt but seriously, fucking DeWalt. Them and Stihl: just gorgeous power tools)

[matwood]

Haha. Yeah. My wife doesn't get it yet that when I pay a little extra for DeWalt it means I will not have to buy the same tool again for a very long time.

[jugg1es]

I wholeheartedly agree with this statement.

[adventured]

No question. The value of sales engines - driving sales referrals - as ecommerce in the US doubles over the next ~10 years (and in many developed economies), is extraordinary. If you want to make a billion dollars, build the next generation of sale referral monsters (RetailMeNot or Coupons.com being a primitive, unsophisticated first generation version; Groupon & Co were/are also primitive early sales referral engines).

A catalog is a concept, rather than a pile of printed paper as perhaps most people would think of it (ie thinking that catalogs died out with the rise of the Web). Historically catalogs pushed sales in all sorts of ways for all sorts of things, for eg the last 150 years in the US (and much longer elsewhere). As a concept, it's a sales referral system; it can either be internally owned (Sears Catalog) to drive content within eg a retailer's selection, or it can drive sales for external stray objects (whether tchotchkes or otherwise). The catalog business in the US was massive for a century. It's being rebuilt online right now. Pinterest, Twitter, Facebook, Snapchat, Instagram, etc. are in part sales referral engines, there will be dozens of major platforms that perform that role, as it spreads to fill in every possible ecommerce niche.

All of these shops/brands/products coming online or being started from day one online, need a way to drive sales online (the more cost effectively the better). Taking a cut of that sales referral action will be dramatically more valuable than rage clicks for content on a random buzzfeed article or a paid click over to low value content on boredpanda and similar.

[hickeygareth]

Are your thoughts that Facebook et al would - instead of charging for an ad 'impression' - charge ecommerce sites a sales commission for the referral? (Disclaimer; I think/imagine they do this already to some extent, but it's not the main revenue driver.)

If this were the case, then it would absolutely be in Facebook's interest to present ads based on a persons likelihood of purchase (perhaps using a deep neural recommendations network (akin to YouTube's) to power it all).

[vl]

Of course they will make more money from this change:

It turns out that using ML to optimize for immediate engagement has two unintended side-effects: 1) it produces junkier content, 2) it decreases long-term retention. For obvious reasons, building a model to optimize for the long-term engagement is way harder and takes way more time.

While in the long run new model is more profitable (due to increased retention life-long engagement goes up), it decreases immediate engagement metrics. When this happens, major accounts start to call in and ask why now they are getting less for their dollar, thus this preemptive explanation by Mr Zuckerberg.

Then, when the dust settles and prices adjust, increased retention will compound and profits will go up.

[xefer]

How many people are aware that the list of interests directing Facebook ads are easily visible and editable?

Settings | Ads | Your Interests

It's really interesting to see the full scope of the interests FB has gathered over time. Some of them are pretty hilarious.

They have a seemingly huge ontology of every subject you could think of. If you methodically go through an remove every interest, the ads suddenly become very generic - stuff targeted to, say, age group and/or location. Since removing everything, and periodically clearing it all out, I generally only see stuff for things like real estate and car dealerships, which don't really mean much for me.

[macNchz]

I found Facebook’s idea of what my hobbies are to be particularly amusing: https://m.imgur.com/nWCWn63

It’s unclear where they got those from, or why those are even allowed to be considered hobbies.

I don’t use Facebook much these days, but I’ve had an account for nearly 13 years, I use Instagram regularly and they surely have lots of tracking pixel data on me...so I was suprised at how poorly they’d inferred my interests (the other categories were less farcical but not especially accurate).

[jp555]

Generally people only consider ads that “don’t mean much to me” as ads. If they do mean something to someone they suddenly cease being ads to that person. It’s a weird thing.

[dhoulb]

I did the same and filled it with things that are relevant to me, and now my ads are generally enjoyable — things I might not have otherwise known about.

There was a lot of cruft in there from the early days when it was easy to like everything. But after the cleanout my ads are definitely better.

[adventured]

> they'll accept that they will make less money as a result of this change

If they're providing less engagement to content discovery from businesses in the feed, it increases the value of the traditional advertising on the platform. Perhaps their plan is expected to simultaneously push up the ad rates they can charge to access users, as an offset.

It's also likely to increase the value of the content Facebook is going to curate/push on its own platform via Watch. They're going to build out a substantial streaming business in the coming years, rivaling YouTube. The user obviously has finite time, this will probably ultimately shift more engagement time to Watch, less time to stray business or media content.

Overall it strikes me as a classic later stage of platform evolution. First you have an ecosystem with large numbers of external parties that are deriving immense value from the platform (whether businesses or developers or other). Then you eat the ecosystem, replacing it with your own systems and on-platform content. For example, instead of promoting a Craigslist (eBay, Poshmark, whatever) post, if you're Facebook you promote on-platform "marketplace" listings. Instead of an external YouTube post getting attention, that goes to Watch. By doing that, they technically fulfill their claimed plan (if only in their own opinion). Twitter, Microsoft Windows, eBay, Google, Netflix, nearly all platforms do this aggressively eventually.

[briandear]

> that content that makes us angry, anxious or jealous is a much better driver of clicks than content that makes us happy

Not for me. I've all but abandoned Facebook because I was tired of being angry all the time. It's a wasteland of political hysterics.

[Turing_Machine]

I just routinely block any political page (of any kind) that someone shares, and have unfollowed anyone who feels the need to go on political rants more than once or twice a week (including, sadly, some family members and long-time friends).

I mean, by this point I get it that they like (or don't like) Donald Trump or Hillary Clinton. Message received.

While I won't unfriend someone over politics (unless they start insulting me), I will unfollow them. If they want to waste their own time ranting on Facebook, it's their business, but I've stopped letting them waste mine.

Twitter is, of course, much worse. I don't understand how anyone can believe that Twitter's format is suitable for serious political discussion, but many people apparently do. I quit Twitter a long time ago.

[Viliam1234]

> If they really mean it, they'll accept that they will make less money as a result of this change.

First, it's not sure whether they will actually lose money. The problem with clickbait is that you learn relatively little about the user. Suppose someone clicked on hundreds of articles about why everyone hates Donald Trump... what does this teach about the user? How does it allow you to show them better targeted advertisement? If instead they will click on stuff written by their friends, you will learn what hobbies they share, and then you can sell them things related to the hobbies.

Second, just because Facebook changed a value of some variable quickly overnight, doesn't mean they can't slowly change it back later, if it indeed turns out to be a loss of profit.

[crorella]

They mention this quite explicitly: "As we make these updates, Pages may see their reach, video watch time and referral traffic decrease. The impact will vary from Page to Page, driven by factors including the type of content they produce and how people interact with it."

I think it's bold but good for the long term.

[dragonwriter]

Every Facebook change seems to reduce Page basic reach, probably in order to increase the pressure to pay for additional reach.

[marbletiles]

> The core problem, IMO, is that content that makes us angry, anxious or jealous is a much better driver of clicks than content that makes us happy.

Right ... and if they're not careful here, they'll just go back to how FB was before the brands all moved in: people posting content that made each other angry or anxious.

A FB feed full of people trolling one other and squabbling about politics in long comment threads might look like it's "sparking conversations" and worth promoting in News Feed, but it's fundamentally going to be the same sort of turn-off as any badly-run forum is.

The core core problem is the content incentives: whether for brands or individuals, FB incentivises content that gets interactions, and without moderation of some kind that leads inexorably to trolling/clickbait.

[JasonFruit]

Wow, who are your friends? I regularly interact with people I'm not angry at, even on Facebook.

[marbletiles]

I have a perfectly lovely set of friends, who often post lovely things.

But when I think back to all the “long comments” and “spark conversation” type posts, they’re not lovely. They are politics or other things that spark FB’s equivalent of a flame war.

So my fear is that using those things to indicate these posts should be more prominent is going to be a tricky thing for them to get right.

Getting a computer to decide between “has lots of comments because it’s a fight” and “has lots of comments because it’s useful and interesting” is an interesting challenge.

[kartickv]

A "Was that worth it?" button might help distinguish clickbait or angerbait from genuinely worthwhile and fulfilling information. Say you participated in a conversation on FB. When the conversation ends, you might get a notification from FB the next time to log in asking you whether it was good.

Or a news site could have a Was it worth it? button at the end of each story, to help identify clickbait or otherwise low-quality articles. Rather than measuring how many page views each journalist drives, they might measure how many satisfied readers that article had, and reward journalists who write high-quality articles.

[JasonFruit]

That is a good statement of the problem. It seems like I've read about some pretty effective tone analysis for English, at least, that might be helpful.

[netsharc]

Well, studies have said social media increases anxiety, I suppose because people usually post highlights of their lives, on average people's reaction to those posts is jealousy and feeling like "Why is my life shit?"...

[joelrunyon]

All they're doing is lowering inventory - this means cpc and costs are going to rise.

What they really need to be worried about is how many advertisers they will/won't lose.

[drewmol]

>It would be the right decision in the long term, but the short term will hurt.

It seems there's quite a bit of incentive to target short term gains for many of the relatively new tech companies. When tons of early employees hold incredibly valuable stock options, which tend to be more valuable than their salary's, this creates an organization wide incentive to drive those stock prices as high as possible during their tenure. Those early employees gain position and influence, then use it to influence company culture and direction.

Is assume this contributes to the reason why it's so hard for behemoth companies to be nimble and quickly change direction.

[lenzm]

It's hard for any large organization to be nimble and quickly change direction, eg. the government, military, unions, religions.

If anything, stock options incentivize longer term thinking. Instead of quarter to quarter goals, stock options reward goals years in the future.

[adrianratnapala]

> content that makes us angry, anxious or jealous is a much better driver of clicks than content that makes us happy

This might be true. Or it might be a belief we formed to explain Facebook's behaviour, and we now need to start updating our beliefs.

Anyway, assuming it is true, Facebook still must balance two things: (a) get clicks from your existing users (b) get and retain users. In their halcyon days they could take (b) for granted. Perhaps now the market demands some humility from them?

[gressquel]

yes, but to counterbalance this they will making companies/pages pay to get articles into newsfeed or on frontpage one way or another.

[vl]

As I explained here, they will make more money from this change:

https://news.ycombinator.com/item?id=16142137

[lmm]

Facebook shows us what we click on. My own feed is, if anything, too saccharine, because I pay more attention to happy stuff than politics.

If you want to make the world a better place, take a look at yourself and make a change, and all that.

[apatters]

Anecdotally, the only change I've noticed is that Facebook seems to be showing more ads (almost always for something I'm not interested in) in my news feed.

[amelius]

> content that makes us angry, anxious or jealous is a much better driver of clicks than content that makes us happy

Especially considering there's only a "like" button and not a "dislike" button. It's frustrating!

[efsavage]

There are more options now, including angry and sad reactions.