A TweetDeck App Platform
In the weeks since Twitter Flight, I’ve been thinking a bunch about Jack Dorsey’s pledge to developers, #HelloWorld, and what many developers (and perhaps users) want: the blessing to develop new client apps, and a release of the token limits that allow unlimited users to leverage those apps. Twitter has probably put a lot of thought into this, and had to know that their pledge to (and plea for trust from) the ecosystem would lead to renewed calls for this holy grail. It’s an unlikely outcome, and could lead to a new swell of discontent. Twitter did, however, make good on expanded platform offerings, and even lifted the covers some on its Gnip data products, making them more accessible and fuller-featured.
But what if they offered a path that assuages Twitter’s (likely justified) reasons to own clients, but allowed developers to build into the client experience? What if TweetDeck had an app plug-in platform?
What Is TweetDeck?
TweetDeck is almost entirely an industry-insider application now, leveraged by many media companies and marketers for managing multiple Twitter accounts, organizing streams and searches, and having a powerful “dashboard of now.” Since acquiring it several years ago, it’s gone through some (sometimes controversial) revisions, but remains largely the most powerful free, and most importantly, official, Twitter client.
Its nearest pure-play competitor would be HootSuite, a 3rd-party client that has thrived despite a changing ecosystem, able to be an environment-of-choice for SMBs to growing enterprises. Hootsuite itself has a plug-in platform, allowing two kinds of integrations (more on that later). They make those available to users, some for free, some for pay, in an app directory.
Using TweetDeck
I use TweetDeck every day. I have 47 columns right now. I often see support requests before our team sees them, and news and Tweets about things and from people that matter ahead of others. It allows me to sift through a massive amount of information in a more orderly, long term way. There’s a search column, with some filtering, on every company we partner with, for Twitter Lists I’ve curated or subscribed to. For events and hashtags that are relevant. For my blogposts and links I want to track. I’m not the only one, and I imagine I’m actually a lighter user than many.
Interesting side note: either because of its 3rd-party roots, or its insider-niche, Twitter Ads do not appear in TweetDeck. Power-users like me likely have one of the most detailed interest graphs Twitter has access to.
An App Ecosystem
This isn’t such an outlandish proposition; Dataminr has an authorized TweetDeck integration. This means that there are indeed entrypoints into the TweetDeck application for outside companies to embed, and they have the appetite to do it with a (very close) partner. TweetDeck today is actually a Chrome Application, running in either a browser or a protected Chrome instance. This could hinder what I’m proposing here.
Twitter today manages a massive application ecosystem ranging from marketing automation, display, logging, analysis and analytics, a handful of surviving clients, and hosts more subtle integrations for publishing, reading, and understanding tweets and Twitter users. There’s a very vocal contingent of developers who feel that Twitter should allow more 3rd-party clients and remove the restrictions that make them untenable. This actually makes sense from a few perspectives:
- The Web 2.0 “Hello World” for introductory programming texts was, for a time, “Build A Twitter Client.”
- Many attribute Twitter’s early success to developers embracing and making clients, and inventing actions, workflows, and [non-developer] idioms, like the @mention, #hashtag, 1/TweetStorm, and $cashtag.
Twitter is sure to have done a heap of analysis here around the actual usage of these applications, and the tradeoffs of user growth with or without them. I still think they made the right decision there. It doesn’t help that the media latches onto this action, perpetuating essentially a myth that Twitter turned their back on developers. I assure you all of us could still create applications and still innovate creatively. Lazy reporting, and in some cases, lazy developing.
One thing missing, though, is a way to curate and leverage the creativity of the developers and businesses building on Twitter. Sure, they have their Official Partner program — and it’s very well run — but there’s no “install this into your Twitter Experience” button. An App Ecosystem, built into TweetDeck provides a way for the features and products to easily be adopted in a controlled environment. It will essentially allow those developers who wanted to build a client to use a client’s foundation and add the features and functionality they’ve dreamed of(within limits) to Twitter.
It might be an open sandbox or a walled garden; it would satisfy developer needs, innovation, and, user growth, if it was tied to a killer feature. That part of the equation is a little tougher to justify. Could Twitter Moments have existed as a 3rd-party plugin? Almost certainly — many companies and publications have performed this sort of curation — but the context has generally been less universal, more brand-oriented. It was never as easy as “side-load this into a Tweetdeck Column.” Now it could be.
TweetDeck With Apps
What would this look like? Hootsuite provides us with an interesting blueprint that can easily be applied (the first two), but we will want to take it further.
Custom Columns
- Applications that process Twitter data, like Nuzzel, could publish their stream of relevant publisher content shared on Twitter, individualized for the user directly in a column.
Custom Tweet and Column Actions
- Maybe you want to easily share a Tweet with another application. Perhaps you’re curating Tweets and, while Custom Timelines (oops, Collections now) are a way to group these together, many applications have their own workflow for this.
- Show related content to a tweet based on a quick NLP analysis
Custom Profile Enrichments
- Klout and other companies have valuable context about who people are and what they are known for and interested in. It’s not perfect, but this is valuable context that many marketers use.
Custom Filtering Rules
- In TweetDeck, columns can be filtered in a strangely imperfect replica of Twitter Search. Imagine being able to have apps offer custom filtering criteria. It’s lazy, but I’ll use Klout here again: “In this column, show me Tweets by people interested in Audi, Volkswagen.” Today, you’d only see Tweets that have those keywords in them.
Potpourri: Custom Notification Rules, People Search on more attributes, Automated Twitter Card creation, Ads API integrations
Getting There
The engineering, product management, and marketing feat to bring this together would not be trivial. Who is the audience for a more full-featured TweetDeck? I imagine it’s still people who use or listen to Twitter as part of their job. We’re all consumers, though; we’re all reading some piece of Twitter that isn’t exclusively related to work; we’re all buying products, participating in social activities.
There’s an interplay here between the three pillars of Twitter’s 3rd-party ecosystem: data from Gnip powers insights that can be surfaced in TweetDeck. Fabric can deliver the developer experience to connect the applications. The value chain supports itself.
Would this satisfy the developer community? Would an aspiring Twitter Client developer find that their ideas were actually untenable? If built into TweetDeck, they’d be given access to an active user base, hungry for efficacy, context, and insights, if not a new consumer context.
