App.Net (its popular nomenclature ADN is used hereafter, mostly) has been through a rough old time in the past few weeks, since Dalton Caldwell's State of the Union in May. In keeping with Maslow's Hierarchy of Needs the worries about the future of such a fine service have really shaken the generally relaxed and adult atmosphere on the service. This short blog post is not about this recent behaviour, which has - at times - been lacking in compassion, understanding and moderation. Instead I want to talk about my personal take on ADN itself, why I joined and why I think there is still a place for such a service. I'll also talk about where I think things went wrong, and what could happen next.
My thoughts about ADN, and my experience with it, fall into 5 broad categories, which I'll address in order:
- Developer respect
- Social Graph
- Quality of Discussion
One of the big drivers leading to the creation of ADN, back in the second half of 2012 was the changing business practices and attitudes of Twitter. The company, in seeking to further monetise its user-base (more on this later) was taking steps to disenfranchise those third-party developers who had been largely responsible for its success up to that point, and was very much reducing the ability for new entrants to this space with (in some cases) fairly draconian changes in the terms applying to new versions of its API (Application Programming Interface - how computers communicate with each other).
This atmosphere was a significant factor in the subsequent developer interest in ADN, along with (I'm given to understand) some incredibly sophisticated APIs and underlying functionality. This generated an upswell of well designed and built 'micro-blogging' clients in the first wave, taking their lead from Alpha, a proof of concept delivered by the ADN team. These clients spanned OS and desktop/mobile platforms and can be found in the ADN Directory. There were some 'world class' mobile apps, in particular, developing rapidly alongside a quickly expanding set of APIs, which would go on to include the capability for Private Messaging, Chat Rooms and more.
Following this 'first wave' was a smaller, but arguably more innovative, selection of apps (both web and native) that leveraged the file storage associated with an account, along with the ADN authorisation and authentication functionality (more below). Unfortunately a shrinking regular user-base (in part due to the lack of an overall compelling brand / product), along with the withdrawal of a 'Developer Incentive Program' - which provided a revenue-sharing 'top-up' for the most popular and useful applications - has meant that there is currently little commercial incentive to continue developing for the ADN platform
As mentioned above, Twitter's business model has moved into the advertising space, using the data held on its many millions of users, along with the content that they choose to share using the network. This has led to Twitter's priorities shifting away from a 'user as customer' model to a 'user as product' model. This can be reasonable, if one is careful to measure the cost/benefit analysis. Does the value we get from the service match the cost, in either attention or the sale of one's data? There is also the matter of ongoing trust - today's business model is not necessarily tomorrow's, and given the recent IPO the fate of Twitter is even less certain than ever before.
There are two elements to this, which are related but distinct.
Firstly there is the fundamental problem of authentication, of proving 'who I am'. Today we have 10-100s of username/password combinations, and while we might use the same username on different services to provide the illusion of continuity (please don't use the same passwords too!!) this is still a fragmented solution, and one ripe for a better approach.
The second is the 'social graph' itself - a 'contact list' of people who you have some form of relationship with. This covers both active and passive relationships, and arguably even one-way (Twitter's Lists, for example), and acts as a 'web' of who we know, and who they in turn have some form of relationship with. Make no mistake, this is of huge value to Twitter and other data-driven companies (such as Google or Facebook), but there's a problem. Intra-company fragmentation has rendered Twitter-provided, Facebook driven and Google services broadly useless - and this also relates back to ownership.
This is MY data - MY relationships, MY friendships, MY interests, MY news sources etc. While this data might be stored on Twitter, Facebook or Google's services (and therefore available to help them drive advertising traffic), this is something that is mine, about me as a person, and their unwillingness to make their services universal mean that incomplete, fragmented social graphs litter the internet - a smidge on Instagram, a little on Twitter, more on Facebook. Like a Venn Diagram made up of broken shards of glass.
The hope of ADN, as a paid (me as customer) service is that the support for this social graph is purely down to developers making use of the appropriate APIs, and is therefore freed from much of the intra-company politics that threaten to cripple the internet of today and the future. This means that my social graph on Alpha (the micro-blogging side of ADN) is also available for developers looking at other applications (such as the Private Messaging application Whisper, or the Chat Room service Patter) - and, in principle, any other future applications.
I don't make light of how hard this is. The issues with the ADN business model make this clear, and indeed the typical user of the internet isn't really aware of the problem, but its one that we desperately need to solve in the coming years.
Quality of Discussion
In the English language (the only one in which I'm fluent), the difference between 140 characters and 256 might seem trivial, but from my personal experience the difference is dramatic - whether making it easier to have a conversation, or to publish one of my haiku poems (never mind the 2048 limit, for Private Messages!). In practice it massively curtails the interactivity on Twitter, leading to something that better resembles people shouting between mountain tops, rather than neighbours talking over a fence.
I wasn't around as a Twitter user in the early days (c2007), but I understand that there was a real sense of community amongst the users, one that's been eroded over the years, as the service has continued to grow at pace. Spammers, trolls, bots, stolen accounts, the news (and Twitter streams) are full of the problems - not to mention the vast number of people who use the relative anonymity of the internet to excuse some of the worst behaviour. Death threats, rape threats, they are out there, and are utterly unacceptable.
The beginning of ADN was a wonderful experience: meeting and making new friends, all of who were willing to pay for a service where their data was respected, and where the sense of optimism was infectious. This included establishing the tone of the service, where a wide range of conversations and debates across the political, religious, cultural and technological spectrum could be had, without the usual descent into a partisan spiral of closed minds and open mouths. This lasted, arguably, through until the introduction of free accounts, necessary to extend the use of ADN as the authentication and social graph infrastructure it was destined to be, but which fatally wounded the community and atmosphere of the service.
From a personal perspective, If I take nothing else from my time on ADN I will take a small number of long-term friendships with me, built on shared values of respect, humour, curiosity and integrity...
What went wrong?
Technically, very little. Where mistakes have been made (with the benefit of 20/20 hindsight) probably fall into 3 main categories:
- Compelling Proposition
- Features & Benefits of Membership Tiers
In summary, the branding of the service wasn't compelling in terms of the wider world - the App.Net brand was too technical, too nebulous given the lack of a mature ecosystem surrounding it. More, perhaps, could have been done to both promote the service more widely, and to directly fund the development of a number of compelling products (think photo management, or journalling, or recipe management etc...) Either way (and I'm sure there are lessons to be learnt that I've not considered here) after a successful launch the service failed to capitalise on the early momentum. There were also a number of wasted opportunities to capitalise on wider privacy concerns, and the exploitation of user data.
Perhaps the final nail in the coffin of the original business model, however, was the roll-out of the Free Membership Tier. At a time when the primary income stream was people using the Alpha micro-blogging service (and supporting third-party apps), this influx contributed to a disruption to the atmosphere of the service, inviting an influx of trolls, spammers and other such undesirable elements. It also enabled some of the service's paying users to make a commercial decision to revert to a free account, while retaining (for them) all their existing functionality. The collective 'shall I renew' posts also served to significantly disrupt the community.
The need for a free account level to support wider use cases (i.e. non micro-blogging) was absolutely essential in support of both those services and the longer-term need to allow ADN to act as a hub for authentication and the social graph. The extension of fundamentally unimpacted micro-blogging access was a significant strategic and tactical error however - one from which ADN may never recover.
All is not lost, all is not ended.
I've given this some thought, and while the situation is serious there are some steps that could be taken.
- Take steps to restrict the micro-blogging abilities of free accounts created after the introduction of 'non-invite' free accounts. This should impact on a limited subset of genuine humans who use the service on a day to day basis.
- Re-brand, with a focus on authentication and the social graph, and the privacy of the user data. A free first year, and a modest ongoing cost thereafter, could make this sustainable. The additional services that are available, or which could come later, are a layer on top of this core feature.
- Some consideration is needed as to additional tiers / paid features - this would allow for a modest cost for the core features, and additional 'add-ons' for storage, micro-blogging, etc...
- Better community engagement. This widens the pool of ongoing ideas, provides an engaged and (still) enthusiastic feedback mechanism, and would reward user advocacy.
It may well be that ADN as it currently exists is past the point where these suggestions could turn the platform around, but any future service would do well to give them consideration...