Facebook increased satisfaction and usage by sending fewer notifications
medium.com
medium.com
That also meant on many days I'd have no notifications at all. A week could go by with 0 notifications. Since the quality of notifications used to be so high, I used to treat FB notifications the same as I treat my email inbox (no email goes unseen).
Over time, they started jamming in "fake" notifications like "Your friend posted a new video!" whereas in the past, I'd simply have 0 notifications for the day. After that, it became impractical to keep up with the never ending notifications.
They need to be looking at the quality of their notifications rather than their frequency.
The best thing that any PM with enough clout at fb could do right now is force the org to track a different metric, and move the content consuming side to a different app (because there’s still a lot of growth value in communities, watch, etc.)
Meta might come up with different revenue streams and different strategies, but for the blue app that ship has sailed.
FWIW, in my time at FB from 2019-2021, we were explicitly disallowed from using time on app as a success metric for the products my team worked on. Instead, we focused on the utility of the experience as the goal, i.e. people finding the product useful was more important than simply using it.
I wasn't working on the newsfeed though, so I'm not sure if that applied to that part of the app.
> force the org to track a different metric
Like adsClicked / serverResourcesConsumed? Interesting how that would stand up for users that e.g. post text vs users that post video, and the fallout of their followers.On another note, I stopped using Twitter recently for obvious reasons. I check it occasionally to see if there's anyone new that moved to Mastodon. And I noticed an increase in fake notifications there. Another reason to never go back there either.
I started using Twitter again and in order to kill fake notifications, you have to go to "Push Notifications" settings and uncheck everything except for "Related to you and your Tweets" section.
Twitter seems to respect those settings, while Facebook still finds an excuse to push fake notifications even with tweaked settings. I basically disabled notifications entirely via OS for FB.
But official reddit app and website in other hand, generate notification badge WITHOUT NOTIFICATION. What did you even expect me to do about that? It's just dumb-ass.
In the end, I just wrote a userscript to remove notification badge of reddit web page, given how useless and distractive it is.
Once I switch to "Inbox" I can see the new message. Of course, if it's a long thread I have to find the bottom of it first.
[1] due to this issue: https://www.reddit.com/r/bugs/comments/rj0u77/reddit_redesig...
Later they came with their timeline which omitted stuff I wanted to see (updates from my friends) and added stupid posts from randos. That's really what forced me off the platform. It broke the whole point of it.
Unfortunately Instagram is doing the same now with their TikTok style "suggested posts".
It would be nice if meta went back to the original inspiration of a “face book” and made a product that was just a photo or two and links to social links and contact information.
After that, it never really got any better, for me at least. Notifications tended to be less relevant than before and I disabled them on all devices.
Then the demographic that advertisers really care about — affluent Americans in the 25-49 age bracket — stopped using Facebook.com every waking minute, and product managers panicked.
Their solution was to add more content to one of the most-clicked areas of the UI (I bet at least one product manager got a nice stock bump based on that change).
Short-term gains are great and all, but the change smelled of desperation, and had utterly predictable long-term effects.
And now we get a blog post telling us what should have been obvious to begin with.
It’s really a shame because I don’t think we’ll see many opportunities in history for an app that everyone has and that can be used to connect people.
I actually tried to figure out how to get rid of this at some point, but it definitely wasn't a priority for FB to give me control over this.
I'll never forget the open contempt Meta has shown for me as a user of their products.
Yeah. They're an advertising company. They make money by spamming users with noise. Contempt for other human beings is in their DNA.
They don't want you to disable the red dot. They want it there 24 hours a day bothering you until you give in and watch those ads disguised as videos just to get rid of it and find some peace. They'll actually run human experiments on their users without their informed consent just to find out the exact level of annoyance they'll tolerate without deleting their accounts in disgust.
If you try to disable the red dot, they'll ban you. It's against their terms of service to not be bothered by notifications. You need to submit to this abuse if you want or need to use their services.
You can still try anyway. Make a plugin to delete the dot. Then a few days later you'll find they scrambled the page structure on purpose just so you could no longer interface with it, precisely so you could not defeat their little nagging dot.
"Contempt" doesn't quite do it justice. They delight in our suffering and profit off it.
That isn't how any of this works.
Even better was when messenger split off from fb proper into its own app, but still showed a notification bubble with red exclamation marks (or some such, it's been a while) that you couldn't dismiss. clicking opened the page to install the messenger app.
I deleted my account.
So not only hostile but incompetent.
A broken feature on the mobile version nudges towards app installation, a broken feature on the desktop version does not.
This means that there is some benefit for Facebook to not fix the mobile version.
If a bug persists and benefits the company, we should - by principle - never assume it to be a bug.
We see the same right now with reddit, where old interfaces slowly (and clearly intentionally) crumble away, permanently signalling to the user that it is time to switch to one of the interfaces with a higher ad density.
Another meta problem is that there seem to be no authoritative product person that can force his/her vision on the product so you get hundreds of visions colliding and fighting for changes. IMO it’s much more valuable to have a coherent vision being enforced for the product, otherwise you get this frankenstein-like collection of apps.
Finally, the fact that HN is such a hater of facebook means that most fb engineers have deserted HN and they’re not reading the excellent ideas and feedback that this community sometimes provide. It’s a shame, because I still completely believe in Facebook as a good product for the world.
Well, people believe the craziest things so you are in good company.
But on the important point on whether facebook is salvageable with a user subscription model: extremely dubious. In some sense adtech has ruined the concept of users paying for a good product and there is poetic justice in that. The magic wand that turns users into product wont turn product into client.
What might work is "social as public or semi-public infrastructure", namely getting cities, companies, states or even countries paying for user accounts of residents, employees etc. as a sort of benefit.
But even in the unlikely scenario facebook would relinquish its residual adtech earnings in favor of a new and less lucrative business model it would likely face stiff competition from newcomers with less historical baggage
So it seems my company is doing that (it's called "workplace from meta" now) and it's ridiculous as you get the worst of both worlds.
I assume people who say that haven’t had a good experience with fb, but it changed my life when I was younger. I still use to track all my friends across the world.
> In some sense adtech has ruined the concept of users paying for a good product and there is poetic justice in that
Twitter blue
I was looking through it, and it started scrolling on its own. Dozens and dozens of stories flashed before me.
Something is really odd, FB has lost a fair bit of my online time recently.
I think you may be right. Just let me pay a few bucks and just show me my friends.
Scrolling through Facebook is like walking through an Outlet Mall in a foreign country.
Which is understandable, because FB optimizes for clickability and shows very few posts of those who rarely post (and whose posts really deserve attention imo) so after seeing no reaction from their friends ("are they ignoring me?" - no, FB just decided not to include your post in their feeds) they just gave up.
facebook is an ad business. to that end, they do try to improve the user experience wherever necessary to advance the goals stated above, but that's it.
https://www.forbes.com/sites/kashmirhill/2014/06/28/facebook...
Subscriptions work when either the user makes money from pushing the content, or actively wants the content they’re provided. On the former, boosted spam has already been implemented, ultra targeted ads are another form of this, and stalking users to push private messages to them is already Linkedin’s bread and butter.
On the latter, the issue is of course that the “content” (people’s post) isn’t and won’t be worth paying for in general. I won’t be paying to read my relatives bickering about stupid issues. Random people’s funny takes and viral jokes will be spread on other networks so paying for it won’t make sense either. I’ll subscribe to my favorite novel, manga or tv show, or check a well put newsletter or podcast. None of that has a place on the blue app.
Facebook at this point has a very specific niche, and trying to get out of that niche either puts it in competition with better/less toxic/griftier services, or canibalizes other meta products. In that sense pushing facebook in a completely new direction is a tall order. If a PM had a crazy novel idea, they’d better execute it in Instagram or a completely new product than have it get beaten to death in facebook.
> haters
I was following a bunch of very talented and idea prolific designers and engineers working at facebook. IMO they haven’t left HN or Twitter or the public place. They left facebook instead.
Or were laid off :))
But more seriously I still believe in fb. I still know a lot of great people there.
I reported the spurious notifications, then eventually gave up and disabled notifications for the entire app at the OS level.
Then seemingly at random it starts showing me short-form sort-core porn videos.
Instagram used to be about fun and creativity, now it's a trashy tabloid that's entirely made of Page 3 content.
For instance mine will show friends stuff until this message, then it starts showing me the random trash.
I'm guessing they're going for engagement (random content > already seen content) but it's always my cue to close the tab, quite handy
You can always return the contempt by extracting all our data out (provided that's important to you) and permanently shutting down you account. I believe that's the most efficient thing you can do to tell FB to go to hell.
Edit: I honestly believe that any FB user this is the most harmful thing you can do to them.
Quitting Meta means not getting to be in those circles anymore.
Basically you're subject to the lowest common denominator.
Message contents also aren't the only kind of data that Whatapp knows about. Facebook doesn't need message contents to monetize your communications. They almost certainly harvest Whatapp's metadata and tie it to everything else they know about you.
If you like the service, that's great. Use it and enjoy. But don't assume that your communications are secure.
WhatsApp encryption is strong enough to defeat criminal investigations in my country. That has to count for something. They got so angry about it they blocked WhatsApp services for days and jailed Facebook people.
https://www.forbes.com/sites/abigailtracy/2016/05/02/whatsap...
A few years back I changed the e-mail to a fake e-mail. I accidentally typo’d my mobile phone number due to which it now resolves with a Philippines prefix. This prevents me from verifying their ‘looks like you’re trying to do a suspicious log-in’
They won’t give access or delete my account despite:
- me knowing the current password
- providing old passwords
- indicating to them the old e-mail address to my account
- verifying that I have access to that e-mail account
- verifying that I own the ‘old’ phone number to the account with a phone bill (!)
- indicating I am willing to provide a scanned ID
- indicating two friend on my ‘following’ list that would be willing to verify its me trying to access the account
At this point I have pretty much done 5-factor authentication and they still won’t accede.
Thankfully I am an EU citizen and the data commissions are currently extremely trigger happy with GDPR fines, so my hope is this forced them to resolve it, or make it cost them tens to hundreds of millions for not doing so.
Fuck you and your scummy illegal practices, Meta :)
And as a revenue-generating set of eyeballs they can monetize and sell to said video content creators.
1. https://www.fbpurity.com/news/important-news-facebooks-legal...
Facebook's blue app is one of the first apps I had to turned off the notification entirely. If I really care about some one who just upload a photo I will go into your app to take a look. These notifications are now just baits for the ads.
When the private sector consistently refuses to self regulate bad behavior, the law can and should step in.
For example, I have notifications enabled for Amazon because I need to know when a delivery is coming but they’ll use that channel to spam me about “deals”. There’s no reasonable way to disable all these ad-via-notifications.
They certainly don’t want you to disable deal notifications and make it very difficult to do so.
It’s like the old quote about fire extinguishers: if you don’t know where one is, it doesn’t exist. If a reasonable effort cannot find the setting, it doesn’t exist.
I'm saying irrelevant stuff like recommendations should never be sent. If I click yes to the above question I should only get relevant notifications.
One thing a lot of people outside the company probably don't realize is that every team at Facebook is different and the culture of experimentation varies greatly from team to team. When I was there, the people who worked on stories ranking were very serious about user satisfaction and thought deeply about whether changes were good for the user even beyond what could be measured via metrics. I worked a tiny bit on notifications experiments at Instagram (wasn't on a team but monitored and changed some ongoing experiments to improve metrics). At the time, I believe the notifications team was less thoughtful about this. I think this has changed in the years since I left though, based on this blog post and what I've seen in the app.
Glad to see someone is thinking about qualitative user experience and correlating it with long-term metrics.
Source: ran an engineering group at a "data driven" consumer internet company that fell victim to these tactics.
I am pretty happy browsers added support for this. It is disabled by default and easy to say no forever if a site is annoying about it.
Even if I disabled pretty much everything, I regularly get a new useless kind of notification on my phone.
If I want to know about a new email, group message or some other random thing I'll open the app or site and take a look.
Otherwise leave me the ** alone.
I for one stopped using email push ony smartphone when my hosting provider started sending out invoices at 1am. Who thinks that this is a good idea?
In the early days of Android there was an app which was able to set my phone to silent based on nearly unlimited inputs, like location, time, app notification, weather, etc. The relevant API was patched out with Android 5 or so, one of the most stupid decisions ever.
It's more challenging to disable notifications we don't care about anymore with tactics like these.
Same with Twitter (though I found tweetdeck solves much of this).
Same with Instagram, trying to shovel content I didn't ask for in my face.
Same with Youtube's dark pattern of trying to get you watch "Shorts" and if you close the row it says "Okay, I won't show you again for 30 days". "NO! FUCK YOU! I didn't ask you to show me again in 30 days!!" (and BTW I'm paying)
Twitter does has that dark pattern as well. Delete something and it says "ok, we'll show you less of this" .. "NO! FUCK YOU. Don't show me any ever!"
The result is that I barely use any of these services. Once a month or so I try again for some reason, boredom? and get pissed off and leave.
Now some FB/Twitter/Youtube person will claim their metrics say that their dark patterns get people to use the service more. Ugh!
This was the nail in the coffin for me. I just want to see what my friends post in the order they post it. But Facebook mangled the feed so bad they drove me away as a user. They’re doing the same thing to Instagram. While it doesn’t show your friends acitivty, it also barely shows anything they post. One post from a friend. One suggested post. One ad. One suggested post. Another ad. Then another post from a friend. Then when you’re “all caught up” it’s just ads and suggested posts. Yes you can tap the logo then tap following to get a chronological feed but I want that to be the default.
The "increased satisfaction" is likely a result of fb going overboard with notifications a few years ago. That phase led me to set aside time one day and actually go through all of fb's privacy settings to turn off all email comms or anything that could prod me to log back into the site. I also uninstalled the mobile app and my Android phone battery thanked me for it.
The best feature by far is the notification dot with the count when you're not (fully) logged-in in the browser. I only ever log in when there's really something that needs my attention. For the last couple of years, my feed is basically births, deaths, marriages - a high no. of posts are by people of a senior age, and the conversation in general just isn't relevant to me.
If only Apple had made reporting of such problems transparent I would have reported this.
They still send me notifications, including notifications that I have notifications disabled.
I don't think I ever get spotify notifications currently. I probably did the above. (Android)
The perception of increased app usage at the expense of satisfaction with the app, was always going to be only short term.
At that point it takes someone with both vision and decision making authority to say say.. nope.. They have to be specific about which ones are too much.
Now imagine that this is true as well to get access to whatsapp or instagram user base. I think whatsapp has really done a good job of not bloating the app and rejecting all these requests.
The ones that don’t use the app yet find it outrageous that other people find utility in it.
The ones that use the app and notice the move away from a useful tool to a Frankenstein app with no coherence.
facebook has enforced the prosecution of womens rights in america and elevated platforms of hate speech and seditious rhetoric in numerous countries. it played a quantifiable role in the january coup in the united states, and continues to promulgate american foreign policy through direct ties with the us government.
facebook maintains round the clock surveillance of nearly everything you do through your phone once the app is installed, and has faced little consequence for it.
The question of if removing notifications that were predicted to be rated 4 out of 5 is not so obvious.
>rather than our usual operating model of sending all relevant notifications (either rated 4 or 5 on a scale of 1–5)
Just because something is "relevant" to me doesn't mean I want to be notified. Usually I don't.
I go out of my way to have zero unread notifications in any software I actively use, even if it means making liberal use of Mark All As Read. This lets me visually scan for where activity is without visual noise.
By definition a relevant notification is something you want to get a notification for.
What might be "relevant" to most people will not be relevant to me.
Imagine needing to spend tens or hundreds of millions to have an answer every single user could give you for free.
You got to love the age of information and social media, gives so many people jobs.
No, the notifications they were sending were low quality.
Think of YouTube. Usually a comment on a video sends you a notification. This may make sense for small creators, but if a creator is getting thousands of comments they may not want to actually be notified of each one. YouTube has to decided if the notification for a comment on a creator's video is relevant enough for them to be sent an actual notification for it.
Oh, they know. They definitely know. What they didn't know is the ratio "(engagement + monetisation)/useful notifications"