Twitter’s multibillion-dollar mistake happened five years ago
mathewingram.com
mathewingram.com
I was a very late adopter of Twitter because I always thought that Twitter and Facebook were the same. They are not. Twitter is slowly replacing my RSS and news reader. Twitter is not a community thing, it's about broadcasting (in real-time, if you need it, most people don't, they are just addicted). That's why I don't tweet and I'm just following 11 accounts. I don't even follow friends, I just follow who has something to broadcast. Twitter is great for that.
Facebook is more noisy by concept. Their focus in communities and friends makes broadcasting side noisy. I had to turn off almost all my friends/likes wall notifications, I don't have the time or will to read that one friend that I meet once a year went to the restaurant or played a game. It's just too much noise. That's why FB is nowadays mostly a chat/email friends app (and FB knows it: they bought whatsapp).
Encouraging everyone to make constant, short, pithy, public pronouncements is not a recipe for thoughtful conversation.
I have a friend who switched to Twitter a few years ago and whenever the fact that I don't use it comes up, I mention that I just can't deal with the inevitable dumbing-down and loss of context that comes with such an idiotic limitation. It's a medium designed for soundbites instead of consideration and "wit"[1] instead of intelligence. His go-to explanation is that the limitations force people to be concise and choose their words carefully. Either he's rationalizing or he's extremely naive about the quality of discourse on the Internet.
[1] I mean that in the most negative way possible. "Perceived wit" might be a better term to use here: I've often seen things like pleasing sentence structure used to cover a lack of veracity, a la "If you're not part of the solution, you're part of the problem". or "YesAllWomen". While not inherently incorrect, I've seen both of those used a massive number of times in situations where they're entirely irrelevant or inaccurate. And yet, because it comes across as pithy and clever, they dominate the conversation.
twitter can be a noisy firehose, and facebook can be delightfully refined.
take a look at two of my interest lists on facebook.
http://i.imgur.com/LRK7Wi9.png (the left column is this list http://i.imgur.com/BCkOXCN.png)
if you have noise problems it s because you follow noisy sources, which is true of both platforms.
http://techmeme.com/river http://mediagazer.com/river http://hckrnews.com
Ive also reduced my skimming vs reading, I just dont actually learn as much. If I skim all day, and you asked me what I read, 90% of it is forgotten. Skimming headlines is like a continual huff that never gets you anywhere. Longform is the way to go.
http://longform.org/ http://longreads.com/ https://readability.com/topreads/ https://www.instapaper.com/daily http://medium.com
Imagine if e-mail was owned by HP. (In this alternate universe, someone at Digital invented it. Digital > Compaq > HP) All email accounts are created at hp.com, everyone's email address is janedoe@hp.com, or bobspizza4344@hp.com, etc. If you pay $240 a year for Enterprise E-mail®, then you can customize the domain name, but hp.com still handles all the sending, receiving, and storage. You don't have a say in that.
Other companies have their own versions of "E-mail", variously called "MS Mail Plus for Windows Live 365™" or "iMail™", but you can't send message between them, so they're only used for niche internal communications. Everyone has an account at hp.com. You have to if you want to talk to anybody.
Some people tried to make an e-mail client once, but HP revoked their API key, so it doesn't work anymore. You have to use HP's e-mail client.
This is how Twitter is (and instant messaging as well, which is a shame. I thought XMPP was going to be our savior in that front, but Google saw to that). For me it's not so much a moral issue, but a practical one. E-mail is vastly better because it's open and interoperable. The WWW is way better than AOL and Prodigy and Compuserve were.
Ultimately, majority of people do not care whether the underlying platform is closed or open - they just want a product that works fine (eg. Facebook). So one of the rare ways open platforms can win over closed ones is if someone invests a lot in an open platform and offers a product with quality on par to those with closed platforms.
But it seems money drives a lot more effort in aggregate[1] than some desire to make the world a better place. So a vast majority of times, we see better products coming out of "entities" (corporations, people) motivated by money, which will do all it can to protect it's money source. Closing the platform seems a perfectly logical thing to do in that case.
[1] I define it as total man-hours spent on something.
Out of all the companies that had a large IM network, the _last_ one to stop s2s federation was the one that "saw to" the failure of XMPP? Never mind the fact that no other company with market power in this space (Microsoft w Skype, Facebook w Chat) allowed federation, walling off 90% of XMPP users, and by the time that Google shut off s2s, it was used by barely anyone.
It's ignorance like this that incentivizes the exact behavior you're complaining about. Companies like Apple and Facebook look at Google's attempts at compromising product quality for openness[1] (which, to be clear, is a trade-off that I personally think is worth the benefits to the ecosystem) and the huge backlash they get when they finally throw in the towel long after the battle is lost (s2s federation was more or less irrelevant by the time they removed it because the vast majority of chat usage was on walled gardens). It becomes clear that the only winning move when subject to the ignorant flailings of public opinion is to aggressively avoid lack of support for things like federation and interoperability from the start, in which case the only people who lose are the users.
Well, that's not logical thinking. Google did try—more than anyone else—to bring openness to IM. When they did eventually throw in the towel, they became the unjustified focal point of justified anger.
Meanwhile, iMessage, BBM, AIM, ICQ, MSM, Yahoo, WhatsApp, etc. all continue running walled-gardem IM systems. The closest thing to an open IM system is SMS.
Thanks for the correction.
It reminds me of a friend of mine who has decided to preemptively leave tech early in her career because of the reputation it has for sexism (despite not (yet?) experiencing any issues firsthand). The main industry she's considering switching to? _Finance_! As much as most people implicitly know that sexism in finance is WAY worse, it's rarely brought up because it's considered something of a lost cause (for now). And yet, in her ignorance[1], my friend's assumption is that the relative visibility of sexism discussion maps to the relative incidence and severity of sexism in these industries.
[1] The term ignorance here is meant literally: not having experience in either field and being forced to go off of what she reads/hears about.
A decentralised Twitter would have been amazing. I've tried playing with all kinds of services to push notifications to a variety of devices and the only thing I've ever found that I liked was Twitter - but I don't want to share private notifications with public Twitter servers (even if I do mark the account as private)
One day I might actually build a service myself - but I have far too many projects on the go as it is.
> decentralised Twitter would have been amazing [..]
> One day I might actually build a service myself
You can even subscribe to people on Twitter, post to Twitter and still get ATOM, RSS and FoaF out on your end.
See https://mattl.io/replies for a more traditional view.
I'm guessing that filter is keyword based
Long story short: network effects are hard.
And amazingly not profitable.
[0]http://www.theverge.com/2014/10/27/7079277/twitter-q3-2014-e...
As to whether that would have proven less or more successful than their current approach, there's no way to know for certain so all we can do is speculate
Some of the other services discussed here do also offer an RSS and/or Atom feed in addition to their preferred format for message delivery; but in all of those cases they're more than just an RSS feed.
Letting third parties do the hard work of experimenting with user acquisition and engagement, then buying up the best and turning their backs on the rest was a masterful move.
We need an open alternative to Twitter, but as a publicly-traded for-profit organization, Twitter, Inc. will never deliver that.
Like GNUSocial and Pumpio? What's your handle?
From my own perspective zot (RedMatrix) should be the future. Whether the socialwg agrees is yet to be seen.Though I have a feeling zot will be the betamax to socialwg's eventual vhs
Unfortunately Twitter has a near complete monopoly in the microblogging market.
Like App.net?[1][2] It would seem you need much more than just being open and "letting third parties do the hard work of experimenting with user acquisition and engagement" to succeed in this space.
This is nothing new, and another tale of caution for people building companies that entirely rely on another company's good graces and policies.
It would have been if you assume that Twitter could thrive on its own. Since they have historically relied on third parties to push their product forward, and then cut them off at the knee, they should have been prepared for the inevitable stagnation that followed, and perhaps they weren't.
The masterful move would have been to carefully augment the base functionality offered by Twitter while encouraging the third-party ecosystem to continue to grow and thrive.
You can see, for example, Apple do this with iOS (albeit at times with a ham-fisted approach) by incorporating some—but by no means all—of the features that third parties come up with into the base OS.
This encourages third parties to continue to participate in the ecosystem while gently nudging them towards innovating instead of stagnating, because that's the only way they will be able to compete with the “free” Apple-provided software and make money.
Twitter was in the somewhat unfortunate position that it was born as a wide-open system that had little in the way of walls, leaving the new management with few options to gain more control over their product but to start restricting what could be done with it.
Instead of recognizing the value of third-parties and commoditizing their complements, though, they went all-out and systematically alienated anyone who wanted to share in their good fortunes. They placed their focus on attracting celebrities and brands (remember when Ev was “excited” that Oprah was joining?) and attempted to create a friendly environment for the average consumer by tightly controlling the user experience.
In itself, I don't think that this is a bad idea—Facebook has done it quite successfully, for example. Twitter's problem is simply that they are about to run out of innovation to acquire because there are no more small timers left who are willing and capable to take a risk on their platform and come up with some unexpected new feature that could later be integrated into product proper.
I think it's a bit of a knee-jerk to start assessing their whole strategy on that basis. It pains me to say so as a developer who would have loved the old platform to remain (I had to shut down a somewhat popular hobby website because I didn't want to put in the effort to switch to OAuth). And as a user who misses a proper TweetDeck equivalent on the phone.
They didn't win any friends in the developer community, but the article is about growth and it's not clear they would have achieved the same level of growth with the majority of users on 3rd party apps.
Shutting down 3rd parties helped them get users back on the native platform, buoying their short term results. However, without the 3rd party developers the native platform is becoming less and less interesting.
We are in a very hot tech market right now so valuations are a little out of whack. Facebook has had fairly strong results the past few quarters which give investors hope Twitter can do the same. We will see!
This is pretty clear from their API policies.
This makes me think that all this locking is not about monetization. It's about protecting their user base.
Facebook is following the same direction. Their current api version does not allow to build an alternative Facebook client.
The transition is one from its origins as a primitive status updater ("What are you doing?") that existed on the desktop and barely on the phone, to what it should have been all along: Snapchat, WhatsApp, etc.
There was a further b.s. step in there for Twitter, when it was temporarily called a micro-blogging service.
The original idea for Twitter was not too distant from what WhatsApp & Co. are today. The problem for Twitter, was that it came along too early, had to compromise on its communication vision (the now massive smart phone market wasn't primed yet), and now finds itself in the quicksand of being a half-breed legacy product that isn't the best solution for any particular thing.
Twitter was supposed to be a communication platform. It failed to become that, WhatsApp is going to have a billion users soon and took Twitter's future, and Twitter has been left behind as an advanced variation of a status update service. It lost its future, and the product as is has a limited potential that will soon be all tapped out.
Only Twitter acquired Gnip can keep using it whilst access has been cut for companies like Datasift.
http://www.theregister.co.uk/2015/04/11/twitter_cuts_off_fir...
http://blog.datasift.com/2015/04/11/twitter-ends-its-partner...
https://blog.gnip.com/twitter-data-ecosystem/
I'm not in a position to say if this is as bad as the 3rd party rugpull a few years ago, as it may only mean people have to pay Gnip for the same data instead.
http://blog.datasift.com/2015/04/21/processing-twitter-data-...
The writing has been on the wall for that for quite a while - there's really no reason for Twitter to sell data to a third-party, who can then sell that data to other people.
End-users of the data are still free to purchase the data from Gnip/Twitter and do whatever they want (within the TOS of Twitter)
Having a third party handle it, take a cut, etc. only make sense if you want to keep customers at a distance (say, to pretend to be concerned about privacy), or if you're too busy or unsure about building the service yourself
Cutting Datasift off (really, just letting the contract expire) is nothing like terminating API access for third parties.
About the main point of the post, I remember not having a good feeling at Twitter's strangling the 3rd party apps. It's possible that they already have all the people willing to chat in public about anything now (and the marketers) so they won't grow much more than this, but for sure a thousand companies developing business models on the top of a 140-characters messaging bus can have more ideas about expanding the platform than the only one controlling the bus. Problem: not all those ideas will benefit Twitter, some of them could even be detrimental to it. I think it's a sort of common dilemma, my platform or theirs, my vision or the one of customers? Historically both choices played well or badly. Twitter made their bet, let's wait another couple of years and see.
http://webcache.googleusercontent.com/search?q=cache:8jwimQ6...
He has reposted it on medium: https://medium.com/@mathewi/twitter-s-multi-billion-dollar-m...
Could someone update the URL to that please?
I don't know how up to speed you are with how Twitter was. There was a point in time that they were infamous for their callous "slaying" of 3rd party vendors. I can't say I followed this story rigorously, but even a passive bystander understood Twitter was not about their ecosystem. They were about Twitter.
I don't think it's a hopeless situation, but in my mind the only real path forward is a world where the largest tech companies build up their ecosystem and embrace 3rd party innovation and ingenuity, and not putting their heads into a guillotine when it suits them.
...if this is a fox luring me into it's hen-house, the hen-house sure is comfy and well-featured. Sad to see it diminished in this article, when IMO it's a far far better offering than any other major companies (looking at you Facebook) are putting out.
Twitter's problem per Ben from Stratechery is they can't grow registered users quickly enough, so they're resorting to building ads for unregistered users. The problem is those unregistered users don't monetize as well.
Tellapart built ad products for ecommerce companies. I've worked in that space a bit, and the problem generally isn't registered logged-in users you know a lot about: you can do pretty good recommendations and it's relatively straightforward to advertise to them; it's just that they are a small fraction of your traffic. So tellapart, by necessity, had to get good at advertising to low information drive-by users. That's what twitter needs; not for ecommerce necessarily, but in general.
It's still a shitty business though; if users aren't logged in and sharing information about themselves both directly and via data exhaust, twitter is fighting on more even ground with all other ad companies in the world and not exploiting the prize that no-one else has access to.
While they need devs now, I think they'll get devs back. I've seen people using machine learning on twitter to do some stuff and there are potentials with any large amount data.
But I wouldn't bet the bank on Twitter as a business or a startup. I was interviewed at a start up and their whole business relied on Twitter and creating a third party software with it.
By becoming an ad network! That's the logical conclusion of all such endeavors in the Valley. :-)
The DoD has had a problem with Twitter where they will not onboard to information sharing programs and will not censor certain types of Twitter content. Other partners, such as Facebook and Google, have been very complicit with the DoD. Twitter is willing to delete accounts, but shadowbanning content and dropping content has made them uncomfortable over the years.
http://webcache.googleusercontent.com/search?q=cache:http://...