How App.net can change everything
A caveat: This article represents my current understanding of the planned App.net platform as laid out by its founder Dalton Caldwell through many different sources, as well as my own personal views of its potential future development. It is not definitive.
What is App.net?
App.net is a service dedicated to providing a new infrastructure for social Web applications that will never be funded through ad revenue. It is the brainchild of Dalton Caldwell and Bryan Berg, co-founders of Mixed Media Labs. The vision for App.net was crystallized as an audacious proposal from Dalton after he received an overwhelming response to a post he wrote on what Twitter could have been, itself a response to a blog post from Michael Sippey, the director of consumer product at Twitter.
On July 13, a fundraiser was initiated to raise $500k in 30 days in order to prove the demand for a new way of thinking about Web startups – one where users pay for compelling products rather than being turned into products themselves for advertisers to pay to market to and mine information about. This funding goal was met on August 12 with more than 7,000 backers donating in tiers of $50, $100 and $1000.
How does App.net benefit users?
App.net is fundamentally driven by a desire to empower users and has the potential to establish a new standard of excellence in the treatment of user data and permissions, which will be innate to the core platform upon which App.net developers will build. App.net has publicly declared that their most valuable asset is their users’ trust.
Some of the things I’m most excited about are:
- Users own the content they create and are already able to download a complete archive of their content at any point in time, something that has been a rarity among Web startups. (Twitter has failed to support this for six years despite declaring the intention to do so many times).
- Users will likely have a universal ID that will connect them across any services they use that are running on the App.net platform. This infrastructure would support seamless discovery of friends across services and a robust view of the permissions a user has granted across all the services they use.
- By charging for access to the underlying infrastructure, spam will be heavily disincentivized.
- As an infrastructure company, App.net’s business motivation is to encourage a vibrant ecosystem of applications and novel uses of data. The most interesting social applications we’ll see in the next phase of the Web will be built on App.net.
How does App.net benefit developers?
In the last few years, the rapid acceleration in the creation of startups can be attributed to a great extent to the arrival of cloud-hosted infrastructure like Amazon Web Services and Web application development frameworks like Ruby on Rails and Python/Django.
Amazon Web Services meant that startups no longer had to worry about how to set up and maintain their server infrastructure or how many machines they needed to buy in order to handle spikes in traffic. Instead, they could plug into Amazon’s system, which treats storage space and computational cycles as a utility. It’s like drawing electricity from the wall – you don’t worry about it running out, and you only pay for what you use.
Frameworks like Ruby on Rails provide powerful yet flexible guidelines and core capabilities for programmers to build Web applications with. They free developers from having to think too much about the common issues of developing a Web app so that they can instead focus on the bits that make what they are building unique.
App.net will combine the simplicity of cloud infrastructure with the power of Web frameworks to deliver the best platform for developing social Web applications. Social Web apps are built around concepts like users, posts, connecting, and sharing. App.net will provide a scalable infrastructure and a base model for these concepts upon which startups can innovate without reinventing the same wheels again and again. Developers will spend less time just trying to make their applications functional, so they can have more time to make them unique and useful.
Is App.net vaporware?
Absolutely not. The current infrastructure for App.net is built using the codebase from PicPlz, a photo-sharing service that supported hundreds of thousands of users and tens of millions of API calls monthly. After announcing the new App.net initiative, a brand-new UI was built in two weeks for an alpha service to demonstrate the viability of the platform. In the one week since this alpha was made available to backers of the initiative, over 3,500 users have joined the service and 40,000 messages have been created. In that same period, 13 Web apps, five mobile clients, two browser extensions, and five API libraries for the platform have been released. There is an actively curated list of projects running on the platform.
Is App.net a Twitter clone?
No, it is not. There is definitely a great deal of misunderstanding about this currently. The service that the early backers of the platform have been using this past week, which can be viewed at alpha.app.net, is a testing ground for the capabilities of the platform. It does heavily resemble Twitter. It also hasn’t been given a specific name to distinguish it from the core App.net platform, and this has contributed to the confusion. For clarity’s sake, for the rest of this post, I’m going to refer to the particular network we’ve been playing with this past week as Alpha. Alpha is just one network running on top of the App.net infrastructure, and in the future there should be hundreds if not thousands. Each of these networks will have its own user base and its own apps, browser extensions, etc., but they’ll all share a common infrastructure and many core capabilities. In fact, they will be greatly enhanced by having standardized ways of talking to each other.
Going forward, I do believe Alpha will continue to play a vital role in the success of the App.net platform, which I will discuss in more detail below. The important thing to realize is that App.net’s core business is not Alpha – it is the platform that powers Alpha.
Will App.net be another Diaspora?
It is understandable to equate App.net to Diaspora, but it is not accurate. I believe App.net will succeed where Diaspora has for all intents and purposes failed, for a variety of reasons:
- App.net is not vaporware. Diaspora was funded on an idea and an initial goal of raising $10,000 but was able to raise over $200k due to the tech community’s excitement over a user-controlled and privacy-focused alternative to Facebook. It took three and a half months before any software was released. App.net has exceeded its funding goal of $500k and established a user service and a developer API during the funding period despite an aura of extreme pessimism by the tech community as to its viability, in large part due to the perceived failure of Diaspora.
- App.net has traction. App.net has exceeded 10,000 backers. A third of those users are already actively using Alpha and have contributed more than 40,000 posts. Well-known third-party developers with extensive experience with the Twitter and Facebook APIs are actively developing tools and services for the platform, have already released working products, and are contributing to testing and debugging the App.net API.
- App.net has a business model. From the outset, App.net will be charging users $50 per year for access to Alpha, and developers an additional $50 a year to access the platform API. Moving forward, App.net will model its pricing on running a sustainable business without any ad revenue.
- Dalton and his team have the necessary experience. This to me is by far the most critical factor. Frequently the human factor is lost when discussing the merits and viability of any particular startup. Building software and a business with the scope envisioned by Diaspora and App.net is incredibly challenging. Diaspora was started by four college students with no prior background in running a startup. App.net has a skilled team of 12 led by an entrepreneur who built a service, iMeem, that at its peak had 26 million users. Founders deal with tremendous amounts of stress, as I have learned first hand several times over. Much of Dalton’s motivation for App.net is a reflection of a deep regret over the mistreatment of iMeem’s third-party developers after MySpace acquired it and then shut down its API without warning. It is a testament to him that he was able to bounce back from this as well as cut his losses on PicPlz and continue to innovate. It doesn’t always work out that way. Ilya Zhitomirskiy, one of the four co-founders of Diaspora, suffered from depression and committed suicide at the age of 22. I think it’s absolutely tragic and I wish there was a greater awareness of just how hard the startup founder life is (I was really hesitant to add this last bit about Ilya but it is a significant part of the Diaspora story and shouldn’t go unmentioned).
What is App.net’s business model?
App.net will charge developers for access to the platform. Controversially, App.net also currently charges users of Alpha for access to the network. There is healthy debate going on about the pricing model for App.net, but one thing is absolutely clear: App.net will not run ads on Alpha and will not have an ad-supported revenue model.
As an infrastructure play, I think App.net has a lot of options for how to develop its revenue model. Here’s what I think they should do:
- Charge developers a basic fee for access to the platform and a network for developers only – similar to Alpha (let’s call it Dev).
- Charge applications based on the number of active users they have. This could be tiered or scale linearly. There should be some basic threshold (say five users) that is free, so that purely experimental apps can still flourish.
- Charge applications based on the resources they consume, the same way Amazon Web Services does. This would enable App.net to be a feasible platform for media hosting in addition to messaging.
- Keep Alpha as a paid-access network for the time being in the spirit of lean development. Those who want to be there in the early days because they feel it is worth it can pay to be there (all the current backers are already doing that).
How can App.net prove its infrastructure model?
In order to succeed, I believe App.net needs to show that it can effectively support multiple networks running on the same base infrastructure and data models, while being able to add their own unique attributes. Additionally, App.net needs to show that it can effectively manage permissions for the data that are shared within and across networks. To do this in the near term I suggest that App.net establish a secondary network for the developer tier of backers, which I called Dev above, and proceed to figure out the methodology for having developer user accounts operate concurrently on Alpha and Dev and how they can cross post between them while maintaining the overall privacy of the Dev network (which requires a different level of subscription payment than Alpha).
Why should App.net continue to build the Alpha network and eventually make it free?
As I linked to above, there is a healthy debate occurring as to whether Alpha should be a paid service for all users. I believe that in the long term it should not, but in the short term the status quo is fine as App.net has already proven its ability to gain traction with a paid approach. There is also some indication from Dalton that Alpha may not live much beyond its current form, as App.net transitions to being solely an infrastructure provider and relies on third parties to establish UIs for interacting with the platform. I believe this would be a mistake.
In my mind, continuing to nurture Alpha is a vital element for the success of the platform. There are several reasons for this:
Bootstrapping the universal user ID. One of the biggest potential benefits of the App.net platform is adoption of a universal user ID that would enable your identity to seamlessly move between an unlimited number of Web services running on the platform. The biggest hurdle with this is the initial identity creation, and Alpha can bear the brunt of this burden.
Give people something to understand relative to Twitter and Facebook. The reality is that grasping the full potential of the App.net platform is challenging. Most of the people who initially visit it will be looking for “a better social network.” They should be able to find it here.
Developers will benefit from a reference implementation. Alpha can serve as a showcase for the best ideas occurring in the ecosystem and provide third-party developers something to measure their own efforts against.
Providing discovery for other networks. Providing users with a central place to see posts originating from a wide range of networks running on the App.net platform will benefit everyone.
I personally believe that it will be in the best interests of the platform to eventually transition Alpha to being an entirely free service. There will be many opportunities for paid networks to spring up on the platform that cater to certain industries or hobbies or age groups, but it will be difficult for anyone but the core App.net team to operate a large-scale free network on the platform without significant investor backing, which leads back to the original problems App.net was envisioned to solve. Again, in the short term I think things are fine the way they are, but in the long term I believe it would be a disservice to the world and likely a poor choice for the overall health of the App.net ecosystem to not open the doors to Alpha for everyone, for free.
Won’t the App.net Alpha network compete with third-party developers?
Yes, it will – for discovering and incorporating fundamental ideas that benefit the entire ecosystem. The reality is it is likely impossible for a platform to improve without competing with services that were created to address the platform’s deficiencies. However, given that App.net’s business model centers on providing the best infrastructure for third parties to build upon, improvements to the core platform should always stand to benefit the entire ecosystem. The best thing App.net can do to prevent screwing over developers is to maintain the open discussion that it has so far (and could continue to do on Dev) and to build a visible roadmap so developers know what is coming.
Will advertising be allowed on App.net?
I believe so, yes. A fundamental misunderstanding about the App.net platform so far has been that there will never be any ads running anywhere on the platform. I believe that is incorrect. What is correct is that there will never be any ads run on Alpha or anywhere else to fund the operations of App.net the company. Ads can make sense in content networks, and any such networks running on the App.net platform should be allowed to run them. But they do not make sense to support an infrastructure service, which is at its core what App.net is. If you want a better understanding of the App.net ad thesis, read this interview with Dalton.
Why doesn’t Dalton understand network effects?!
Excuse me, but I believe Dalton understands network effects better than almost anyone in this industry right now. While everyone is up in arms about the chilling effect of gated access to a social network, they are completely missing two other potentially massive sources of network effects: a developer / platform ecosystem that supports thousands of interoperable services and a core infrastructure that provides an extremely high level of customer satisfaction. Not only has Dalton demonstrated profound insight into network effects at a macro level but at a micro level as well. He has stated he intentionally and carefully titrated the addition of new members into Alpha so as to not have an inverse network effect by having a social network filled with n00bs who have no idea what is going on and who end up hating the experience. If you’re looking for network effects on App.net, I suggest you look here.
What are the risks?
Although I have been extremely excited by the potential of the App.net platform, there are certainly many things that could go wrong. Here are a few that come to mind:
- Investors. App.net has been built by Mixed Media Labs using existing code from PicPlz. Mixed Media Labs has already received several million dollars in investor backing and it isn’t immediately clear how in line Dalton’s backers are with his new vision, or how much control he maintains, or essentially what, if anything, prevents the company from being forced to follow the same path that Twitter seems to be on. I feel pretty confident that there isn’t anything to worry about here, but it would be very nice to have some more concrete information from Dalton.
- Baggage. Much of the conversation on Alpha up to this point has been about how to implement “missing” features that can be found on Twitter. There is a real risk that innovation will be stymied by the pursuit of copying existing lousy social-networking mechanisms. However, there have also been some truly great threads that have questioned fundamental assumptions about how these things should work, and again I’m pretty confident that there are great things coming.
- Rushing. Building a robust yet flexible API to support the widest variety of social-network implementations is no easy task. Having lots of users on Alpha has already put pressure on the team to build rapidly, and this could come at the expense of “doing things right” for the long term.
- Security. Having thousands of networks utilizing the same core user ID could be a recipe for disaster if accounts aren’t secure.
- Decentralization. App.net will have to figure out how to provide redundancy or better yet how to decentralize as the platform grows. We don’t want thousands of startups to grind to a halt due to central points of failure. We also don’t want all these startups to cease to function should App.net the company decide to shut down, for whatever reason.
If it succeeds, App.net will undermine the basic economic premise of the entire current social Web ecosystem, and this is a good, good thing. This is a tremendous opportunity to dream big and put aside ingrained thinking.
Editor’s note: To comment on this story, visit Orian’s blog.