As I write this, join.app.net just met our 500K goal, with 38 hours left.
Data Export
When you are logged into the App.net alpha, we provide a button which will email you a .zip file of all of your content in a structured format. If you are an alpha tester, go ahead and try it out. It works.
Impartial 3rd-party verification of results
We are using Stripe to host/power the billing aspects of join.app.net. In the very near future I will ask an impartial 3rd party take a look at our data (while preserving all privacy of our backers) and publicly verify that the join.app.net was operated in an honest manner. There has been zero manipulation of numbers, or “stuffing of the ballot box” by App.net.
Third party app development
We are excited to see quite a few 3rd-party apps already under development. If you are interested in taking a look, here is a crowdsourced directory of 3rd-party apps that are active or under development. This is especially exciting for us given that the API has only been live since Tuesday evening.
This is just an alpha test
Please understand that we built a functional web application and working API to demonstrate that App.net is not “vaporware”. We have a great deal of work to do. One of the most important things we need to do is put together a Terms of Service for the operating site. I will be spending a great deal of time in the coming days creating a draft of our ToS, and our forward plan is to host it on github. This way, folks can see it, offer feedback (even pull requests), and will be kept abreast of any future changes. Along these lines, there are still a great many questions that need to be answered before App.net should be thought of as an operating service, rather than just an alpha prototype.
Account claiming
Please note that once the backing period is over, users will no longer be able to “claim” their Twitter usernames. From that moment forward usernames will be awarded on a first-come first-served basis. We implemented “claiming” as a fringe benefit for our backers, not as a go-forward plan. I want to make sure that latecomers are not surprised and disappointed to see that they can no longer get their preferred username. To repeat: if you want to “claim” your Twitter username, and haven’t already backed us, you have 38 hours left to do so.
To our early backers
Thank you for believing.
I know in my heart that what made join.app.net succeed was your willingness and openness to give App.net the benefit of the doubt, to read our github documentation, to ask to participate in the alpha, to write blogposts in our support. Thank you.
We couldn’t have done it without you.
Brennen Novak just wrote an interesting post that nicely sums up some of the criticisms regarding join.app.net.
At the bottom of his post, he asks me to publicly state/clarify some important questions.
Perhaps that’s biggest flaw with how App.net was presented thus far. Perhaps if Dalton had promised one or more of the following…
I am publicly stating that, if backing is succesful, App.net will support the following things:
- Activitystrea.ms Atom & JSON feeds, as well as RSS feeds, of public posts for individual users, hashtags, etc. (Note that this is different from making them the foundation of our read/write API, which we have decided not to do)
- Pubsubhubbub (PuSH) support (as a publisher, initially)
- Exposing user identities with Webfinger
- Commitment to coordinate between internal and external parties to create and support open-source “lightweight” clients in as many flavors as we can, ala Stripe
- Commit to enabling and supporting users in building inbound and outbound syndication to and from App.net
Long story short, as the social web evolves, we want to be able to provide the hooks into our service for users to syndicate content in and out via current and emerging standards.
Please let the above list show that we will pragmatically implement features and standards at the request of our customers: developers and users.
We shipped an alpha version of App.net. You can browse the global feed at alpha.app.net. This is a webapp that we built in the last two weeks on top of our documented API. Think of this web application as a “proof of concept”.
We also released a dev API that allows 3rd party developers to begin building App.net applications. In the 12 hours since we released the dev API we have already seen several app developers start working on projects. Here is a screenshot of an iOS app that an alpha tester is building. We also have had an alpha tester successfully post messages from an Android app.
We’re opening up the alpha to all backers of join.app.net that want to be included. There are still some rough edges, but feedback has been positive.
Some thoughts on the funding goal
We are currently 43% of the way towards our goal, with 5 days left. A lot of folks think that we will never hit the goal. I disagree.
One of my friends that works at Kickstarter explained to me that projects succeed or fail based on the first and last 24 hours. For that reason, we are not making contingency plans. Rather, we are trying to put ourselves in the best position we can be for the last 24 hours of the project.
Additionally, if you take a look at Kickstarter’s official stats, it would appear that of 35,138 unsuccessful projects, only 2,026 of them ever reached 41% or more of their funding goal. In other words, only 5.7% of Kickstarter projects that don’t succeed ever manage to reach 40% of goal. That is not to say that join.app.net doesn’t need a great deal of support to succeed within the next 5 days, I am simply pointing out that the data would suggest this is not a “lost cause”.
If you have been considering backing App.net, but have stayed on the fence because you are afraid that App.net is “vaporware”, or because you think that backing us is a waste of time (because it will never succeed), I would encourage you to question those assumptions.
Thanks.
Mark,
On June 13, 2012, at 4:30 p.m., I attended a meeting at Facebook HQ in Menlo Park, California. In addition to myself, the meeting was attended by executives at Facebook with the following titles: “VP, Engineering & Products”, “VP, Partnerships”,“VP, Corporate & Business Development”, and “Director, Developer Relations/Open Graph”.
As I understood at the time, the purpose of the meeting was for me to present/demonstrate a new iOS app & service I have been building on the Facebook Platform. Previously, I had been reassured by Facebook dev-relations employees that the service I was building was an interesting/ valuable use of Open Graph & Facebook Platform. I was hoping the outcome of this meeting would be executive-level support for my impending product launch.
The meeting took an odd turn when the individuals in the room explained that the product I was building was competitive with your recently-announced Facebook App Center product. Your executives explained to me that they would hate to have to compete with the “interesting product” I had built, and that since I am a “nice guy with a good reputation” that they wanted to acquire my company to help build App Center.
I quickly became skeptical and explained that I was not interested in an acqui-hire. I said that if Facebook wanted to have a serious conversation about acquiring my team and product, I would entertain the idea. Otherwise, I had zero interest in seeing my product shut down and joining Facebook. I told your team I would rather reboot my company than go down that route.
Strangely, your “platform developer relations” executive made no attempt to defend my position. Rather, he explained that he was recently given ownership of App Center, and that because of new ad units they were building, he was now responsible for over $1B/year in ad revenue. The execs in the room made clear that the success of my product would be an impediment to your ad revenue financial goals, and thus even offering me the chance to be acquired was a noble and kind move on their part.
I am not sure if this bubbled up to you, Mark, but after this all happened I directly communicated my feedback regarding just how unhappy I was with this situation to one of your executives. The executive apologized and said he would take my feedback under consideration.
Mark, I know for a fact that my experience was not an isolated incident. Several other startup founders & Facebook employees have told me that what I experienced was part of a systematic M&A “formula”. Your team doesn’t seem to understand that being “good negotiators” vs implying that you will destroy someone’s business built on your “open platform” are not the same thing. I know all about intimidation-based negotiation tactics: I experienced them for years while dealing with the music industry. Bad-faith negotiations are inexcusable, and I didn’t want to believe your company would stoop this low. My mistake.
In a lot of ways, I got what I deserved. I have come to the conclusion that I took this foolhardy risk because the Twitter “platform” was even more of a joke than the Facebook “platform”. As someone that wants to build quality social software, software that doesn’t force users to re-create their friends list, or not use oAuth, etc., I have to endure huge platform risk. Personally speaking, I am resolved to never write another line of code for rotten-to-the-core “platforms” like Facebook or Twitter. Lesson learned.
Mark, I don’t believe that the humans working at Facebook or Twitter want to do the wrong thing. The problem is, employees at Facebook and Twitter are watching your stock price fall, and that is causing them to freak out. Your company, and Twitter, have demonstrably proven that they are willing to screw with users and 3rd-party developer ecosystems, all in the name of ad-revenue. Once you start down the slippery-slope of messing with developers and users, I don’t have any confidence you will stop.
I believe that future social platforms will behave more like infrastructure, and less like media companies. I believe that a number of smaller, interoperable social platforms with a clear, sustainable business models will usurp you. These future companies will be valued at a small fraction of what Facebook and Twitter currently are. I think that is OK. Platforms are judged by the value generated by their ecosystem, not by the value the platforms directly capture.
I don’t think you or your employees are bad people. I just think you constructed a business that has financial motivations that are not in-line with users & developers. Even if my project isn’t the mechanism that instigates this change, the change will happen.
Mark, based on everything I know about you, I think you get all of this. It’s why you launched FB platform to begin with. Do remember how you used to always refer to Facebook as a “social utility”? That is an interesting term to use. I haven’t heard you use that terminology in a while. I can guess why.
Anyway, Mark, perhaps the public markets & your employees will give you the time and goodwill to fix the obvious structural flaws in your “platform” business. You are in a very challenging position right now. Good luck.
Respectfully,
Dalton Caldwell
Most people don’t remember this, but Google had a social network that predated Google+. In fact, Google’s social network launched a few months before TheFacebook launched as a Harvard-only site.
Google’s social network was called: Orkut. Orkut was invite-only, and the day it launched, an Orkut invite was a hot commodity in silicon valley. I seem to recall I was lucky enough to be one of the first few thousand users.
It was fun to use at first, but within months of launch, the user community started to change. I remember that as it started to really grow, I began getting several new friend requests a day… all written in portuguese. As the community rapidly changed, a lot of early users, including me, stopped using the site. Check out these demographic charts if you think I am exaggerating.
Wait a second, the conventional wisdom is that more users makes a community more valuable? Well, not in this case. US traffic and usage dropped like a rock. Orkut became an embarrassment to Google, and they quietly stopped talking about it. In fact, many people believe that Google stayed out of “the social networking space” for as long as they did because of how poorly they fared with Orkut. I wonder if Facebook would exist in its current form if Google had managed its nascent social strategy differently. Possibly not.
Critical mass
Everyone can agree that an online community has to have critical mass to be considered useful. Getting to critical mass is a high bar: most consumer internet startups fail because no one cares about their product. Many startups are left with a ghost-town, with nary a tumbleweed blowing through. This is one of the reasons that “grow your userbase as fast as possible at all costs” has become VC dogma for consumer Internet. I can absolutely understand the reasoning behind this, but dogma is still dogma.
One additional reason that growth above all else is VC dogma is the assumption that consumer internet companies are free & ad-supported. Think about it: if you have a free service that doesn’t scale a large enough userbase, you end up with a desperate company that cannot financially sustain itself through advertising. Sounds like circular reasoning to me.
The “Suggested Users” mirage
I know a smart reporter that has 429,000 Facebook Subscribers. Really impressive network effects, right? Well, take a look at the comments on one recent news story she posted to Facebook. Here are some of the “likes”. WTF? This is the best her 429,000 subscribers can come up with? Are there other people in her following that are scared to participate? Are her core fans just not seeing her posts in their newsfeed? Something strange is going on.
Similar things happen to other people I know with an abundance of Facebook subscribers. Facebook must have pointed a ton of random users at high-profile influencers, but it seems these random users are profoundly confused/looking for a date. Is this value? Is this what advertisers and Wall Street are looking for? I’m not convinced.
I personally went through a similar experience with Google+. I was added to over 5,000 Circles within weeks of launch. I remember asking my newfound community why they added me. The responses I got suggested these folks had no idea who I was, or how I even ended up in their stream. I suppose it made me feel special to have that much of an audience, but really, aren’t these big numbers a superficial numbers-game more than, well, quality?
Social networking tech press reminds me of The Three Stooges. Pretty much every day we get to read about Moe, Larry, & Curly poking and slapping each other with growth & engagement metrics. It’s entertaining, but the entire discourse is centered on who among them can build the most profitable-seeming vanity metrics from the perspective of Wall Street and advertisers.
Anti-network effects
Anti-network effects occur when a community which has already achieved critical mass begins to lose value with each additional signup. The reason is that the core community that created the value to begin with starts to get marginalized and leaves.
In my opinion, Quora is more “valuable” than Yahoo Answers. But I would argue that Quora could easily become Yahoo answers if, in the pursuit of “network effects”, they begin to dilute the quality of the community, and which would have the side effect of causing the most interesting and value-adding users to vacate. The Quora team is clearly aware of this risk, and are apparently steering the ship in such a way as to avoid this possible outcome.
The power of the asymmetric model + global feed
Twitter’s growth model is a nice blueprint for getting a critical mass and then growing to global scale. Specifically, if you launch with a small, dedicated group of interesting people that can asymmetrically follow each other, along with a global feed of all content posted, you can feel like you are the member of an interesting and vibrant community.
As the site starts to scale, the early userbase will depend less and less on the global feed, and use their own feed/following list to crank up or down the amount of information they are presented with.
The asymmetric follow model also takes care of some of the strange things that happen on Orkut, Facebook, Google+ etc. Strangers can choose to follow you, and @-reply to you, but it doesn’t feel like they are “putting” their troubling messages on your content.
It should also come as no surprise that Pinterest and Instagram followed the Twitter blueprint of asymmetric follows + global feed to scale from a small critical mass of interesting people into a massive, global community. Those sites were fun and useful to early adopters on a small userbase, and have managed to keep their community mostly solid throughout massive growth.
App.net: social infrastructure, not a media company
When I look through the current backers of join.app.net, I am excited to see that it appears to have a large overlap with the early adopters of Quora and Instagram. What these early adopters have in common is an interests in quality, novel approaches to old problems, business models of 3rd-party “indie” developers, and willingness to be a part of something new and strange. I could not imagine a better group of people for join.app.net.
I believe that a critical mass of users and developers can take the basic plumbing we have in our API and webapp, and build a vibrant social ecosystem. My personal hope at this point in time is that our official webapp functions as a proof-of-concept that the API is sound, but is a bootstrapping medium by which novel integrations are built. Remember: Twitter community members, not employees, invented hashtags, retweeting etc.
The point of my post, What Twitter could have been, was to point out that they had the chance to jump off the precipice, abandon their official apps, and let the entire service be the API. Instead, Twitter chickened out and have decided to systematically control and destroy the ecosystem in the name of advertising. I am trying to make join.app.net a “do-over” on this mistake, armed with the benefits of hindsight and a radically different business model that has fundamentally different alignment of user & financial incentives.
If we look to history, infrastructure innovations such as electricity, cellular phone service, and the Internet are judged by the longterm value that was enabled and circulated throughout their ecosystem, not the amount of value that corporations directly captured. If “social platforms” really are as fundamentally important as many of us believe, they should start behaving more like infrastructure, and less like the entertainment industry.
A story
Imagine that there is a hot new restaurant in San Francisco. The restaurant is currently selling hot dogs. However, they insist that there are some brilliant engineers from MIT, Caltech & Stanford in the back room working on a machine that can turn hot dogs into caviar.
This sounds a little bit far-fetched, given that various other restaurants have been trying to build a hot dog-to-caviar converter for the past decade, and it doesn’t seem to have worked yet. But this new restaurant insists that there is another restaurant in Mountain View that figured out how to convert hamburgers into caviar, and that their restaurants have a lot in common. In fact, they even hired some of the same hamburger scientists from Mountain View to work on the hot dog converter in San Francisco.
However, the fact remains that this mythical hot dog technology doesn’t exist yet, and the way the restaurant makes money today is selling hot dogs. Also, the restaurant is under intense financial pressure to get the machine working, and is valued by investors and employees in a way that assumes the hot dog-to-caviar machine already works. They have roughly 12 months to get the machine working or Bad Things will happen.
New social advertising units
The advertising units Twitter and Facebook are selling today are “hot dogs”: poorly targeted, poorly performing. Surrounded by offensive user-generated content that is bad for brands. These are low-end, volume businesses.
Remember all of the ridiculous statements MySpace used to make about “HyperTargeting”? That was their version of a hot dog-to-caviar machine. How did that work out for them?
One of the following statements must be true:
A new “social ad unit” will be created at Facebook or Twitter in the next 12 months that will manage to be far more profitable than current ad units, not piss off users, and immediately be embraced by advertisers.
A new social ad unit will be invented in the next 12 months, will be attractive to advertisers, but will be something that makes end-users run away in droves. Remember Facebook Beacon?
A new social ad unit won’t be invented, and the Facebook/Twitter monetization path will be to cram their current low-value ad units down the throats of users at a far higher volume than they do today, in order to hit revenue targets.
I hope you like hot dogs!
p.s. I am trying to start my own restaurant that transparently sells good food to people that are interested in buying it. No magical machines.
I have been thinking a lot about what makes great 3rd-party developer ecosystems work. At the end of the day, it really boils down to financial incentives, and the ability of an ecosystem to support 3rd-party devs making a living and maintaining a good lifestyle.
If the rules are setup correctly, great 3rd-party development platforms create a strong financial incentive for 3rd-party developers to make great software. Why? Healthy platforms allow 3rd-party developers to make lots of money. If you can setup the financial incentives in the right way, people are able to make a great living by building great software that is useful and makes people happy. That is the world I want to live in.
On the other hand, if the “rules” of a development platform are setup without intention or thoughtfulness, the ecosystem will fail.
I would like to publicly share an idea for App.net and its relationship with 3rd-party developers. What if we took the subscription revenue we got from members, and did a recurring, monthly revenue share with 3rd-party devs?
For the sake of argument, let’s assume a 50/50 split would make sense. In this hypothetical proposal, if in a given month an App.net member signs into a single 3rd-party client, that developer will get $2.08 from App.net. If a member uses 5 different 3rd-party apps that month, the revenue will be split between those 5 3rd-party developers on a pro-rata basis, probably based on how much the user actually used the client. Let me be clear: the definition of use is very complicated, and there will be insanely complicated corner cases to work out in terms of reporting, not least of which are developers attempting to game the rev-share system. Is there a fair way to define “use” that could fairly compensate the ecosystem? Perhaps someone out there has proposal about how this could be fair? I am fully aware that the devil is in the details here, and don’t want to appear to be flippant about a very complicated question.
I like the idea that every month the 3rd-party developers will get a report of what usage occurred in their application, along with a check. The advantage is you get to distribute your app for free in the app store, but still get assured you will get financial incentive for building the best software. And, if a user is already an App.net member, they can just sign into your app for free without any money changing hands… the dev still gets paid at the end of the month. I think this creates an incentive for developer to create amazingly high quality apps that people will want to use in the long run, and get away from some of the App Store “chart gaming” going on today.
Let me be transparent here: I am trying to turn the brain-dead way Twitter is treating its 3rd-party devs on its head. 3rd-party developers added all of the value to the platform in the early days, and instead of being shut down they should get paid. The reason the Twitter developer ecosystem has failed is that Twitter itself has a flawed business model. We can do better.
3rd party devs: what do you think? I think you would say “hey, great idea, but where are the users, this only works at scale”. My answer is that we need to cooperate to create this ecosystem. If we all act together, I think we can build an ecosystem that will support all of us, and lead to you making a whole lot more money building great software than you are today. I believe this is a vision worth fighting for.
Fred Wilson is one the smartest, most genuine people in the tech business. He has a huge fan club that he has earned by being radically transparent and consistently engaging in public debate. I have emailed with him a few times over the years, and actually met him and shook his hand a few weeks ago. I think Fred Wilson is awesome, and if I saw him on the street today I would walk up and shake his hand again. One great thing about people like Fred is that you can have respectful disagreements and debates with them, without it devolving into a non-constructive flamewar.
With that being said, I am now going to make an argument that his blogpost, “In defense of Free”, which was apparently intended as a response to my blogpost, “An audacious proposal”, is wrong.
The semantics
So to start with, let’s talk about semantics. I am making arguments about Web2.0 services. One defining aspect of a Web2.0 service is that it has an API and is, ultimately, a platform.
This word, platform, is a very important one. Companies like Facebook and Twitter have actively encouraged companies to think of their APIs as “platforms”. They want people to base their businesses on top of them. If you unpack that word, they are saying that you should think of the APIs they provide the same way you think of an operating system like Linux, or a hosting platform like Amazon Web Services, or a programming platform like Ruby on Rails.
Platforms are great because they enable you to get all sorts of benefits from the work done by others. I am convinced that Facebook and Twitter really are platforms in that sense of the word, because if you choose to use their APIs you can do amazing things that would be impractical if you attempted to build the entire service yourself from scratch. The concept of platforms is one of the key reasons that Web2.0 actually did meet its promise of widescale global adoption and technical innovation. Long live Web2.0.
However this “platform” word starts to get very troubling when talking about business models. Building on top of a platform is a foundational risk, and if your platform decided one day that it doesn’t like what you are doing, or likes what you are doing so much they want to compete with you, it’s Very Bad. Your platform partner can easily damage your quality of service, or simply shut you down. If that happens, your business is dead. Web2.0 built a lot of really cool, shiny things, but the foundational aspects of them are built on what I am arguing is a flawed premise. I am not simply criticizing, I am saying we can do better.
Fred Wilson makes some arguments about Pay TV. Great. Is he saying TV is a platform? Did HBO build its business and audience by providing an API for app developers to extend… and then systematically acquire or destroy the developers that basically drove their early success? My arguments are that platforms should not be ad-supported. I never said that “free” is the wrong business model for everyone. Of course it’s the right business model in certain cases. How exactly did Fred refute what I actually said?
I would summarize many of his other points as “Dalton’s idea will never work, we have seen people try this before, and it’s a waste of time”. OK, great, we’ll all find out, won’t we? But the title of his post is “In defense of Free”, not “Dalton’s proposal sucks so everyone please stop talking about it”. To mount an effective defense against an argument, it’s usually a good idea to address the offensive argument… right?
On “being a bitch”
I tweeted:
Remember when @fredwilson said that Twitter had no monetization strategy at start & developers shouldn’t build on it? http://techcrunch.com/2011/05/23/fre …
I wanted to remind people that Fred Wilson made the following statements:
Don’t be a Google Bitch, don’t be a Facebook Bitch, and Don’t be a Twitter Bitch. Be your own Bitch.
He also said:
Twitter didn’t have a set strategy at the start. If it were planned, Twitter would have launched with all platform clients.
That strategy sounds like join.app.net, right?
At any rate, Fred replied to my tweet:
@daltonc i never said developers shouldn’t build on twitter. i said “be your own bitch”. that means building for everything, incl Twitter.
I find his response illuminating. He seems to be admitting that Twitter, Facebook and Google treat their platform partners as their “bitch”. In the Fred Wilson view of the world, life as a 3rd party developer is nasty, brutish and short. The best defense is to hedge the companies that will invariably make us their “bitch” by playing them against each other.
And hey, if we’re lucky, maybe they will be kind enough to buy us before they put us out of business.
Fred, I want to live in a world where I am not anyone’s “bitch”. I’m trying to build a platform where the core values and associated business model are set up to ensure no 3rd-party developer will need to hedge against the possibility that, one day, their platform partner will decide to make them their “bitch”.
I think my proposal is a whole lot better than yours, Fred.
The media business
I spent the majority of my 20s building an incredibly large and ultimately unsuccessful digital music business. Because of that experience, many consider me an “expert” on the digital music business. If you have the time to watch my post-mortem on that experience, you will hear that I walked away with the feeling that the human beings that work in the music business all seemed like essentially good people trying to do the right thing. The problem was, their control and scarcity-based business model was/is intractably flawed. The vast majority of innovations created by digital music startups will be crushed because the labels have a fundamental financial incentive to do so. This is the way that I feel about most “media” business models. Attempting to create a huge platform business that is at its core about controlling and monetizing “bits” is a fool’s errand.
It seems very clear that Twitter (the company) is under the impression that the “content” in their system, ie the manifestation of tweets, social graph and “intention stream”, is their core asset. Once you take on that “let’s control our content” state of mind, no matter how fundamentally good your employees, management team, or board members are, you are screwed. Twitter is already well into their harrowing journey down the slippery slope. It seems that most folks in the tech world, except the people still working at Twitter, know how this story is going to end.
Open vs Closed
Fred seems to be implying that a paid business model is more “closed” than the “open” ad-supported one. I think he is dead wrong. Media business models are predicated on control of content. The business model of the join.app.net proposal is predicated on providing a service to its users and developers. In my business model, trying to monetize in and around other people’s “content” is not our reason for existence. The propagation of “content” through our pipes is just the side effect of providing an amazingly useful service. In other words, content is not king– it’s just bits passing through our system, at the behest of our customers.
In my view of the world, digital “services” are valuable/easy to monetize, and digital “content” is not. If we think of Twitter and Facebook as communications platforms, rather than media/entertainment sites, it seems that their business models are on the wrong side of history.
Remember
Whether you agree with me, or disagree with me, I am not just another complainer on the Internet. I’m putting my money where my mouth is.
I’m trying to create the service I want to exist, because I am trying to create the future I want to live in. The alternative is depressing.
On getting public attention
The number one thing that I tell new founders before they launch their product is to remember that no one will care. I have launched several products, dozens of major announcements, and hundreds of incremental improvements. Without exception, the wider world’s response to those launches was smaller than I expected.
It is very rare to be sitting on an idea that manages to capture people’s imagination. My understanding of just how rare having an inherently powerful idea is prompted me to pay a great deal of attention to the amount of discussion this blog post generated. I have never before worked on something that, through no real effort my own, got picked up and propagated in such a wide manner.
Sure enough, when we announced the join.app.net project, it got more attention and inspired more strong opinions than anything I have worked on in my career. For myself and the team, putting something out and immediately seeing the Internet jump all over it is exciting, humbling and scary. All at the same time.
How community, and trust, is built
I have thought a lot about the communities that I trust the most, and my belief is that trust is built through a constant, constructive, bi-directional flow of information.
We are painfully aware of the less-than-perfect launches of projects like Diaspora and WakeMate. Both of those projects seemed really interesting to me at first, and unfortunately didn’t manage to win my long-term trust. We will do everything within our power to not repeat that same pattern.
We announced this project early, with the explicit goal of having a collaborative, open conversation with the external world. We expect to be having a constant conversation with the community, asking for feedback, figuring out who is interested in what we have to say. Listening. Learning.
How we will run our API development process
We are going to be running our API documentation and development process in a fairly public manner. We will be posting a draft of our developer API to Github this week. We will be accepting pull requests.
When you see our API documentation, it will be fairly straightforward. You won’t see anything that will be totally unexpected, or anything so new and amazing that it will blow your mind. You may see things you don’t like, or see mistakes that we made. But this is just another step in our two-way collaboration with the community.
I spent a few years in the early 2000s lurking on the Linux Kernel Mailing list. Having read a great number of the threads and flamewars, I am convinced the success of Linux is very much tied to Linus Torvald’s approach. Overall, he comes across as a kind, deeply thoughtful and patient man. On some topics he is very open minded… he seems very willing to seriously consider deeply weird proposals, as long as the code attached to the patch is good. On the other hand, he does not suffer fools and is completely unwilling to engage with assholes. Linus is willing to defend his ideas and not just push people around by fiat because Linux is his baby. Most of the time, he trusts and delegates to his lieutenants in the community, but occasionally he will ignore everyone and do what he thinks is the Right Thing.
To be clear: App.net is not running a completely Open Source project, and I don’t claim that myself or anyone on my team has 1/1000 of the talent and intelligence consistently displayed by Linus Torvalds. But I bring up his development approach as a model of what a wildly successful open development process looks like.
I deliver to you a promise that we will be constantly communicating what is going on. To listen to anyone that provides constructive criticism, and that wants to engage in providing advice and suggestions on systems they have built. However, please note: if you are an asshole, we will ignore you.
We will make mistakes
A lot of folks didn’t like the way we implemented our Twitter username claiming process. Some helpful folks started directly telling us they didn’t like it, and we had a public conversation about it. Within a couple of hours we rolled out some new code with their explicit suggestions, and haven’t received a single complaint about it since. There were a few “Internet Celebrities” making fun of us for this, and that’s understandable. But I would ask: doesn’t our behavior of listening and fixing the issues raised by the community demonstrate that we are following our core values?
These kinds of mistakes will happen again. Maybe next time they will be harder to fix, or take more time. All I can say is that we are listening, and if folks want to constructively push back and suggest a better way to do things, we will do our best. There is no chance we will be able to make everyone happy, but I want to believe we will be able to convince folks that we are part of a two-way conversation. That’s what I expect out of the companies I love, and so I want to hold my company to the same standard.
I want to hear from the Builders and Makers
In the past few days I have met with Ralph Meijer, who wrote the Jaiku side of the (unreleased) code that would have federated Jaiku and Twitter. He was very helpful and provided a great deal of context for me. I am going to be meeting up with Brett Slakin, the co-creator of PubSubhubbub in the next few days. I am going to be sitting down with Tim O'Reilly in the next week to get his thoughts and ideas. I have been talking to a few other people over email, but they have requested that I not make public that they are engaging with App.net at this time. Given the delicate political position they are in, I understand their concerns.
If you are a builder or maker that has suggestions or pushback for how App.net should work, I sincerely want to hear from you. I just ask that you are willing to do so in a constructive, concise and unemotional manner. If you are willing to hold up that end of the bargain, then I give you my word that I will listen to what you have to say, and do my best to integrate your thoughts into our vision of App.net.
If you like what we have to say, and want to ensure that this project gets to see the light of day, please back us here. If you remain skeptical and are still on the fence, I understand. All that I can humbly ask is that you follow our project updates, and keep enough of an open mind for us to potentially earn your trust in the future. Thanks.
The overwhelmingly positive response to my blogpost, What Twitter could have been has been inspiring. The post has generated 80K pageviews thus far. Without really meaning to, I touched a nerve.
The responses to my post largely fell into two camps. One group is of the belief that a non-commercial, open source, open standards federation of real-time protocols is the solution. The opposing group has pointed out that these decentralized efforts never work out, and the API-focused service I wish existed is the fevered dream of navel-gazing geeks. No, these people say, we must swallow our bitter ad-supported medicine… it’s the only way.
I think there is another option.
A relevant story
My first programming job was at a company which was then called VA Software, working on a product called SourceForge. At that time, If you wanted to launch, manage, collaborate on and distribute an Open Source project, you used SourceForge. The problem was, SourceForge was an ad-ridden, user-hostile piece of crap. Getting anything done required several extraneous pageviews & clicks. The site was designed to squeeze every last advertising penny out of you. I can’t blame management for trying to generate more revenue…. two months into my job there, 25% of the company was laid off.
There was much public hand-wringing over the crappiness of the SourceForge user experience. There were two camps in these debates: those that wanted to build an open source, decentralized version of SourceForge (which someone did), and those that pointed out it was a free service, and how dare anyone complain about the user & developer-hostile aspects of the experience. Tolerating the bad behavior of SourceForge “is a necessary evil”, the apologists would say, “otherwise the service we all depend on might go away.” Does any of this sound familiar?
Years later a site called Github came out. It was good. They had no advertising, but charged money for certain features. They quickly became profitable because the service was so good and so important, people were willing to pay. Github has become a much-loved brand and service, and many would agree that it is a key piece of infrastructure in the technical renaissance we are currently experiencing. Github is apparently profitable, and it sounds like the people that work there spend their time trying to make the best service possible, as opposed to spending their time trying to extract additional pennies out of their users.
Advertising is not the Only Way
Github and SourceForge were both based on providing a hosting platform + collaborative services built upon Open Source tools and documented, open protocols. (Git & CVS respectively). Because they are open source, anyone is free to use CVS/Git without having to adopt the centralized services offered by for-profit companies. But I deeply believe having for-profit, centralized companies innovating and operating these services is a Very Good Thing. Additionally, I think it’s clear that a paid, profitable organization as the steward of the service is far superior to its ad-supported counterpart.
It’s also worth noting that Github succeeded not because the SourceForge team wasn’t exceptionally smart and talented, and not because they didn’t see or understand the user-hostile moves they were making. SourceForge’s lousiness was not because venture capital, or centralization is inherently evil. No, I think the takeaway here is that the services provided by SourceForge/Github are too important to its users to be ad-supported.
Contemplate for a moment how scary a theoretical purely ad-supported Dropbox would be. I can easily imagine the overly-cheerful corporate blogpost explaining why placing ads in my personal documents, or selling the file-listing of my music collection to the music industry, or shutting down IFTTT API access is “important to the health and welfare of the community.” I happily pay to avoid that nightmare scenario, wouldn’t you?
The advertising-supported monoculture
“The best minds of my generation are thinking about how to make people click ads. That sucks.” - Jeff Hammerbacher, fmr. Manager of Facebook Data Team, founder of Cloudera
As consumers, we are currently given the choice between Facebook, Twitter, or Google+. Oh, and there are also some startups with hand-wavy future advertising business models. All of these services are essentially in the same business: vying for the opportunity to sell you/your clickstream to advertisers.
Why isn’t there an opportunity to pay money to get an ad-free feed from a company where the product is something you pay for, not, well, you. To be clear: I’m glad there are ad-supported options, but why does that seem like the only option? For example, I have the option of buying a Mac if I don’t want to buy a crapware-infested PC, right? I have no interest in completely opting-out of the social web. But please, I want a real alternative to advertising hell… I would gladly pay for a service that treats me better.
The disappointment of Web 2.0
I was a part of the Web2.0 movement. I have attended Foo Camp. I attended Bar Camp. I attended Tag Camp. A full-page picture of me was included in a 2006 Newsweek cover story about Web2.0. One session at Foo Camp this year felt like a wake for Web2.0. We discussed the progression: a free service with a vague business model captures the hearts and minds of a large user base, and becomes vitally important. Because the hosting bills and payroll balloon as the service grows, founders are left with a very difficult decision to make. Sell the company? Cram the site full of ads? Keep raising money to delay having to deal with this issue as long as possible?
How do I dare to call all of this out? Well… because I was founder/CEO of one of the huge Web2.0 disappointments. I ran a company called imeem, which users loved, but which I was unable to keep going because I couldn’t turn a profit. I raised >$50MM for that company, had 100 employees, and managed to sell $2MM of advertising per month at peak. There is not a day that passes that I am not filled with remorse and regret for all of the users (and employees) that I let down. I’m sorry.
After imeem, I created a photosharing service called picplz, and made monetization a first-class focus. Fairly early on I gave up and spun that product off, one reason being that it became clear to me that the space was following the same old Web2.0 cycle. If I am honest with myself, I just don’t have the will to ever play that particular game again. Running a service that is important to people with some sort of hand-wavy platitudes about someday launching a “new form of advertising” is not OK. It’s not fair to users, it’s not fair to employees, and we should all know better. This all boils down to a fundamental issue: if an online service doesn’t have the trust of its users and developers, then what does it have?
What I have been building at App.net
I have been working on a service called App.net for the past year. App.net is a paid service for mobile application developers. In addition to the publicly available App.net developer tools, we have been spending the majority of our engineering resources the past 8 months building a “secret project”.
This “secret project” consists of a consumer-facing iOS app & service. During this development process, we have spent a great deal of time thinking about realtime feeds, developer APIs, and creating a service that we enjoy using. It’s also worth mentioning that this iOS is the highest-quality piece of software my team has ever built. We have learned a lot of native app development lessons the hard way… but tough lessons ultimately yield fantastic software.
And now, my audacious proposal
I believe so deeply in the importance of having a financially sustainable realtime feed API & service that I am going to refocus App.net to become exactly that. I have the experience, vision, infrastructure and team to do it. Additionally, we already have much of this built: a polished native iOS app, a robust technical infrastructure currently capable of handing ~200MM API calls per day with no code changes, and a developer-facing API provisioning, documentation and analytics system. This isn’t vaporware.
To manifest this grand vision, we are officially launching a Kickstarter-esque campaign. We will only accept money for this financially sustainable, ad-free service if we hit what I believe is critical mass. I am defining minimum critical mass as $500,000, which is roughly equivalent to ~10,000 backers.
Are 10,000 backers really a critical mass? I think so. Although Paul Graham is specifically describing a hypothetical new search engine rather than a new realtime feed service/API in this inspiring blogpost, his assertions about the power of 10,000 committed users are highly relevant:
The way to win here is to build the search engine all the hackers use. A search engine whose users consisted of the top 10,000 hackers and no one else would be in a very powerful position despite its small size, just as Google was when it was that size.
Since anyone capable of starting this company is one of those 10,000 hackers, the route is straightforward: make the search engine you yourself want. Feel free to make it excessively hackerish. Anything that gets you those 10,000 users is ipso facto good.
In Closing
Many people will criticize me, my proposal, my past failures, my motivations. Fair enough. But I care too much, and think this is too important to run the risk of not trying. I knowingly risk a huge public failure because I truly, sincerely believe in this project.
You, the people that read my blog, tweet my posts, and comment on Hacker News have fired me up so much about this proposed service that I have had trouble sleeping. Thank you for inspiring me. I know in my gut this proposal can and should succeed, but I need your help. Please help me manifest the service that we wish existed.
Concrete details about this proposal, and how to back it, are here.
