Instagram to increase ad load as Meta fights revenue decline
techcrunch.com
techcrunch.com
In optimization algorithms like stochastic gradient descent, some form of entropy (crucially temporarily decreasing fitness) or exploration (having parallel evaluations of the fitness function) is necessary to not get stuck. What mechanism incentivizes this in a large corporation, where every initiative must aim to improve the business and there is the capability to actually measure it all?
A Visionary CEO would chase the mission and expect revenue to follow.
If that was ever the case. In words of Mark Zuckerberg "dumb fucks".
"They trust me — dumb fucks," says Zuckerberg in one of the instant messages, first published by former Valleywag Nicholas Carlson at Silicon Alley Insider, and now confirmed by Zuckerberg himself in Jose Antonio Vargas's New Yorker piece. Zuckerberg now tells Vargas, "I think I've grown and learned a lot" since those instant messages.
https://www.gawker.com/5636765/facebook-ceo-admits-to-callin...
> When I asked Zuckerberg about the IMs that have already been published online, and that I have also obtained and confirmed, he said that he “absolutely” regretted them.
It doesn't say that Zuckerberg confirmed the particular quotation that you offered.
That said, I take your point. The New Yorker author said that he had confirmed the message containing the quote, and Zuckerberg didn't deny sending the message, and he expressed regret about it.
[1] https://www.newyorker.com/magazine/2010/09/20/the-face-of-fa...
We'll find out in 10 years if Mark Zuckerberg was a visionary genius or an idiot.
If Meta continues destroying users’ faith in Instagram the way they have in Facebook, it will gravely hurt any attempt by Meta to build their AR/VR empires.
A good counter example is Apple, a company which actually realized how impactful this can be with their Mac business.
Apple’s Mac business is their second smallest, ahead only of the iPad (which thanks to shared development costs and components with iOS is almost certainly more profitable) and behind accessories, services and iPhone.
And yet, Apple invested significant R&D, and spent years publicly apologizing for the state of the mac, humiliatingly reversing course in some decisions and promising they would do better until they actually did better.
Because Apple recognized that their commitment to the Mac meant their customers could trust them. And the years they were ignoring it (probably in a misguided belief that the relatively small size of the segment meant it didn’t matter) led to a massive loss of trust that was clearly spilling over into their iOS products as well.
Merely having statistics won't lead you to local optima, it's the incentive structure. Public companies need to have a positive second derivative on their revenue, which trickles down to PMs and directors doing anything in their power to improve the quarterly numbers at long-term costs.
I don't think that's true. In fact I don't think very many public companies have accelerating revenue.
Which law of physics is this the result of?
Note that “reasonable” may not be Buffet’s desired p/e.
All IG/FB can do now is squeeze as much ad revenue as possible while there are still hundreds of millions using it daily.
Spotify is borderline unusable to me now, it's about 50/50 ads/songs. But I still haven't uninstalled it. I think my usage is now something like, after listening to 5 songs and 5 ads, I get frustrated and switch to a podcast. But, I don't know of any alternative with less ads that works on my phone, so I'm still a user (despite listening for an hour a week instead of 40 hours a week).
The current internet media giants got huge by burning VC money for years to operate with few or no ads, and now Insta/Youtube/Spotify are all basically as bad as broadcast TV and radio for ad density. But is anyone going to give a new upstart a billion dollars and 5 years to host ad-free content to establish a new competitor?
Facebook and Instagram? Nope. You can't even pay them for an experience which lets you focus on what you want. Which means that the only economic incentives come from the ad revenue side, and the user engagement side will be seen primarily in terms of manipulation.
Course, we have the family plan, and UK prices, so I get ruined. But now I have Netflix I hate ads with a passion, so there we go.
You should consider paying for the service because what you're doing is self torture.
except with Instagram it's, "this is so much less usable than the last time I tried it. Why does anybody put up with it?"
To be fair people were saying the same thing 20 years ago when AIM, ICQ, MSN Messenger and Yahoo messenger were the big thing. All messaging platforms will eventually die is my experience thus far.
That's generous.
Also, it’s worth noting that HN isn’t exactly representative of the population at large.
This is going to kill engagement. I'd rather pay more for clicks than now be jammed into a crowded feed for hostile eyeballs.
Hopefully they really think about this before doing something so stupid
And now, the era of social media is over. Instagram's death leaves Twitter as the last relevant social network, and Twitter's prospects don't look too rosy these days either.
The fourth era of the Internet appears to be content distribution platforms serving a large audience that passively consumes the content of (or, at best, interacts parasocially with) a small cadre of creators. The digital frontier has been fully settled and it now looks an awful lot like the old country.
Given the rise of TikTok-style endless scrolling through short video content, there is a surprisingly apt comparison - channel-hopping through hundreds of channels when you know it's all crap.
We went full circle! Maybe really next step will be some more personal and decentralized media.
The real change that drove all those billions in ad money was simple: under no circumstances can the user's scroll experience end due to the inventory of scrollable blocks ("content") running out.
The user doesn't know right away that the software will scroll infinitely. Email has an end, your IMs have an end. With social media, you have to tell yourself it's time to stop. Most people still won't do that until they reach a state of boredom.
The machine, having learned that, is tossing away old-fashioned images and text, and is switching to video to go even further to distract you, in hopes of keeping you for a few seconds longer, so you can see relevant advertising about some Wish.com gizmo.
The people who post these comments will always come back and post them. And the usual war will ensue.
After all you can argue vim v BSD and GPL v Emacs for absolutely free on ad free sites and people always did. Contentiousness rules when contentiousness is supported, as on sites like this.
Or perhaps it's a natural characteristic of social networks.
- You use RES or an old version of Reddit, or access via a third-party app, which doesn't support ad display from Reddit's network
- You don't subscribe to subreddits where Reddit does have many advertisers. Crypto and investment subs in general are crawling with ads. Reddit can't attract the same breadth of advertisers that Google and FB can, so they can only show what ads they have in their inventory.
Increasing ad load from that? I guess it's time to retire my Instagram account.
This might be the thing that gets me to bite the bullet and actually check out Pixelfed.
Funnily enough, Facebook has been showing me ads for lingerie the last few days, although I put in real age and gender (male), and it should know I'm single...
Instagram uses Facebook/Meta algorithms for serving ads, which includes behavior. I almost never post anything about diabetes in Instagram, but because many other sites (ie: reddit) use Facebook libraries for whatever reason they are using them, Facebook knows I'm diabetic, and I get served ads for diabetes products, bogus supplements that "cure" diabetes, health programs that use diabetes products for non-diabetes purposes (yes, this is a thing), and so on.
If a site uses a FB tracker library (which is going to be part of FB based auth), your usage is tracked and info based on that is used to serve ads across all Meta products.
I will never get why FB shows so terrible ads when they supposedly know so much about us.
So Pixelfed wouldn't really change things for me except give me a better experience while shouting into the void.
Tiktok started as a Chinese only service. China has restricted access to the global internet, so having their own video social media platform was a big deal. It built up a huge audience.
When it went global, it had a built-in audience of viewers and creators. Yeah, they may be Chinese, but an American or British or what-have-you outside of China person posting a video on there and watching viewer counts go brrrrr makes you want to do it some more. It attracts other users and creators.
So a combination of timing, exclusivity to a region with restricted global internet, and simplicity in the design makes it attractive. An excellent trojan horse to collect data on a massive number of people around the world.
Why? Facebook's decline doesn't mean social media's decline. They'll just go to TikTok, which I understand is even worse [1].
[1] And by worse, I mean a slicker and more addictive product. It might feel more enjoyable, but that's actually worse, IMHO.
I am considering using ejabbered and wondering what your thoughts are on it today. Thanks
At this point I feel embarrassed to upload anything. Not that it matters, engagement from other people is at an all time low. Comments seem to be mostly bots.
TikTok is an echo chamber of narcissists and people with unwarranted self importance, but dialed to 11, just like twitter, but with more ADHD. Reddit is the embodiment of mediocrity, ruled by wrongthink policing moderators.
I am not enjoying social media anymore, I feel there is no place for me.
It is worthwhile if you can avoid the traps of attractive people dancing, casual outrage politics, cute animals, and people doing stupid things.
Edit: regarding the article adding the new advertisements, I hope they don't mess up some of the reels that are perfectly looped. I recently discovered those and some are incredible. Reminds me of the lions in kenya flash video that looped perfectly.
I agree so much, really dig the analogy :D
For example, I just checked Instagram and got an ad for Amazon. In that ad they were trying to sell me a log splitter, a peanut butter squirrel feeder cage, 25lb sack of peanuts, and a aluminum garbage can you would find in a park. These things aren’t even remotely relevant to me.
XD peak capitalism I guess
For grid posts (not sure what they're called, but the part that's not stories, where you scroll down), my feed is currently as follows:
1) Friend
2) Friend
3) Ad for a prescription medicine I'll never take. Totally irrelevant.
4) Friend
5) Friend
6) Ad for a product I'm unlikely to buy (it's a personal grooming product)
7) Friend
8) Friend
9) Friend
10) Ad for a retail item I'll never buy (item costs $5k+)
11) Friend
12) Friend
13) Friend
14) Ad for a movie that I probably won't see, but this ad isn't totally off base
15) Friend
16) Friend
17) Friend
18) Ad that's not relevant to me (it's for a car)
19) Friend
20) Friend
I also usually see "suggested" posts that aren't relevant, but none showed up in the first 20 posts today.
So...5/20 = 1/4 of the posts on my feed are ads. If they bump this up to more like 1/3, I'll stop using Instagram.
Now I seem to get stuff tailored to pilots - like Airbus vs Boeing jokes, stuff about cockpit controls that mostly goes over my head, etc... I'm not a pilot, and I don't think I'd ever want to fly a plane, but it is a lot nicer as far as what shows up in the feed.
It has no ads, but needs to be killed and reopened for a full feed refresh. I say that's a plus.
They need to rely on the army of data scientists and smart people still left there to help guide their decision making through data. Increasing ad load solves the short term problem of depressed conversion reporting thanks to privacy changes in the Industry.
It DOES NOT solve the business problem of a rising competitor in TikTok, a failing ads business model (for growth that wall street expects) given present privacy challenges.
So FB tried to chase TikTok, and ended up (a) making it much harder to follow and connect with your actual friends, and (b) ended up with a shittier version of TikTok.
Unfortunately, it's not that simple. A lot of people "connect" online by raging about political obsessions. A "no politics" rule will make a lot of people angry, especially if it wasn't like that from the start so violates their expectations.
> It DOES NOT solve the business problem of a rising competitor in TikTok, a failing ads business model (for growth that wall street expects) given present privacy challenges.
You've got it backwards. The "army of data scientists and smart people...guid[ing] their decision making through data" will tell them to make tactical, nonstrategic decisions like this, because they're about measurable stuff.
And that problem was clear long before the ascendance of data-worshiping computer nerds: https://en.wikipedia.org/wiki/McNamara_fallacy.
"We did an A/B test and figured out that we can increase ad revenue by x if we increase amount of ads shown to users by y! We are geniuses!"
I do hope these large firms with such a wealth of smart people on the inside have robust debate on the nuances of what "success" looks like as you say. If someone pushed a formulaic approach of juicing metrics, you need a healthy ecosystem of pushback.
Facebook is caught in the IBM self-fulfilling spiral of death. It's the financial engineering direction as the stock wilts and the multiple is compressed lower; that FB began buying back its overvalued stock, destroying its cash because they had no idea what to do with it, was a big signal.
During October-December 2021 they spent $19.2 billion on stock buybacks. Around 58 million shares. The stock was ~$330-$340 then, now it's $136 (those 58m shares are worth sub $8b now). It's a clown show. Those people have no idea what they're doing.
It makes me wonder, what would it take to run a social network in steady state? It sounds like it would need to be a non-profit or at least private organization. Closest examples that come to mind are Craigslist or Wikipedia.
It's never finished. People learn and adapt to avoid IG's addiction mechanisms. IG then has to zag its approach to get "viewing time" numbers back up.
I would definitely pay for something like above if it meant 0 ads, I despise ads and go out of my way to avoid them when I can.
I just don't get it. If you think further than the next quarter this seems like a terrible long term choice. Can someone explain how a company like Meta arrives at a decision like this?
You are giving this person too much credit. His only other big thing besides getting lucky with Facebook itself is the metaverse, a soulless "Second Life 2 with VR and blockchain support".
> Can someone explain how a company like Meta arrives at a decision like this?
Maybe everyone just gave up, they know they have no way to fix Meta, so they are just gonna squeeze the lemon really, really tight one last time?
To take something like Facebook from very successful college site into the behemoth it is now with luck alone one needs a particularly shining star.
He didn't build 1 trillion dollar company -his team did. He was just there at the right time and did the right thing when he developed the initial version of Facebook, something any somewhat competent full stack developer can do -and I am explicitly talking about the initial version, not the planet scale real time super systems his team later came up with.
I'd argue if any of us were sent back in time just one year before he started working on Facebook, we could easily be the Facebook founder of that timeline.
Mark is famous for keeping decision privileges to himself and for being central managed who's very involved in the business. Are you arguing that the successful results of his decisions are arbitrary? That doesn't make sense - like I said, others tried before him, even got some traction but lost the game. Others tried afterwards, Google for example, and lost. He survived snap. He incorporated Insta and Whatsapp. Do you really think that any average fool with the ship and the money could get some smart guys to sail the seas for him?
The ultimate question is:
There are hundreds of thousands of good dev teams, why don't they all build $1trn companies.
Maybe this was because I was in some long lived experiment group (unlikely) or maybe because the mobile web app doesn’t have ads (also unlikely). I’m curious the reason.
It's already quite disgusting at this stage. Are they going to flip it to 2 ads every one post? How long before unskippable 30-second ads?
They don’t increase ads in feed, but add them to new places like explore.
Only US-centric ones are failing after years of dominance, and even then, most have accumulated enough money to pivot into something even more horrifying.
I just don't see that applying to any of those, certainly not G+ :)
Facebook / IG ad performance dives hard as the number of impressions per ad / per person grows (ie. People scroll by ads they have seen on autopilot).
My guess is they have already accidentally introduced "a bug" which reduces the effectiveness of impression capping but I've got my popcorn ready for the full removal of impression capping.
So given that this is so obvious, why are they doing it? Are they really that desperate?
I know it's a public company, but shouldn't Mark Zuckerberg have enough pull to not run into these short term this-quarter issues?
- 2 posts from people I follow
- 1 ad
- 1 "suggestion"
So half of my feed is not my feed. It's slightly less than a month ago when it was 2 thirds of ads/suggestions.
1 of the ads was also extremely intrusive in that it was related to a private health issue I had a few weeks ago.
Like an erp or something.
What am I missing?
I'm a photographer and just posted a few images from an engagement session. I wanted to tag them for the added engagement. On my personal Facebook page they both pop up on the search right away as they're local to me and in this case we have some mutual friends. On my business page I was only able to tag one of them as he had a unique name - the girl has a common name and Facebook oh-so-helpfully just gives me 5 or so random people from the entire world with her name with super tiny thumbnails to boot.
I decide to try and tag while on my personal account and realize I can't. I know I (and others) were able to in the past. I look for the setting that allows people to tag the images on my page. I can't find it. I look at their support and it mentions a button that clearly isn't there for me. I find a Youtube video from just a few months ago and their interface matches mine but the settings page for my page is completely different and is lacking many options, including the "Tagging ability" one that I need to turn on. I spent an hour on this to no avail, and like many of you I have a degree in computer science and I'm no stranger this.
Another thing I previously spent at least an hour on is the fact that my business Facebook page at the very top shows the amount of people I'm following - 9, and a competitor of mine doesn't. It's a stupid little thing but it's the first thing people see. Alas, I can't figure out how to change it.
On Instagram, when creating an "audience" for ads or boosted posts the option to center who the ad is shown to around a location/town with a specified radius in miles is just straight up broken for me. Doesn't work. I've tried reinstalling the app. The best "workaround" is to either list every single town I want to target in an area or physically drive to where I want it centered on and do it from there. The address input doesn't work but my current location does.
On PC, I can switch between my two business Instagram accounts once before it doesn't work and I need to log out and log back in.
Pushing reels instead of photos may or may not be a good business decision but it absolutely sucks for us photographers and when I tried to make one using their built in tool it demolished the quality of the photos I was using.
Also, ideally I'd be able to target people that are recently engaged to be married instead of people that have been for a while. People usually get a photographer (and venue) quickly, but my only option is to advertise to everyone that's engaged.
Meta - you want more money? Fix your shit. I'd love to give you more money but you make it so incredibly hard.