CEO of Twitter Dick Costolo. Francois G. Durand/Getty Images
Twitter is releasing a new software development kit today at its developer’s conference. It’s called Fabric. It does lots of very pretty things, and the people who write mobile apps are going to eat it up. Which, OK, that’s nice. But the bigger story is what Fabric represents. Because it isn’t just a tool for developers any more than Greek horses were meant to beautify Troy. Fabric is the foundation for Twitter to transform a business based purely on a single product—tweets!—into a diversified service aimed at every person and company that makes mobile apps. That, in turn, would affect every person who uses mobile apps. In other words, everyone.
If Twitter succeeds with this plan, it won’t matter whether or not you use Twitter the product. You will end up using Twitter the company every time you use your phone—even if you’re not aware of it. Up to now, the only way a company could insert itself on such a grand scale was to make a handset or an operating system. But there isn’t going to be a Twitter phone or a Twitter OS. That leaves it with no way to get in on mobile in a big way—unless it lives in applications themselves. Which is exactly the plan.
The key is enticing developers—with some of whom Twitter has had a tortured relationship at times. To do it, Twitter is going to give them—at no charge—a suite of tools designed to help them sign up new users, analyze their applications usage and stability, and, most enticingly, make money. In exchange, the developers will build Twitter right into every app that gets written.
The payoff for Twitter will come if it can get developers to embrace MoPub,, its advertising product, because it gets a cut of any ad revenue. So it gives developers these nice tools to sign people up and improve their apps’ performance—and oh yeah, here is a click-to-install ad plug-in to go along with that, which happens to make us money.
“If we achieve our goals with Fabric it will be the first thing developers include when they write their apps.”
“This is a big step for Twitter,” says Kevin Weil, who runs all of Twitter’s business products and leads the Fabric effort. “It’s Twitter becoming more of a mobile services company. This is us stepping in to play a more powerful role in the mobile ecosystem, based on what we’ve learned from being a mobile-first company for the last seven years. If we achieve our goals with Fabric it will be the first thing developers include when they write their apps.”
Kevin Weil. courtesy Twitter
If that sounds audacious, it is. This is Twitter in transformation. It has been working toward this moment since it acquired a company called Crashlytics in January 2013—a curious move at the time. Why would Twitter want to buy a company that helped developers figure out what was making their apps crash? It followed this a few months later by purchasing a mobile ad network, MoPub. That made more sense, but still wasn’t a part of its core business.
Now, those acquisitions suddenly seem crucial. They form the backbone of Fabric—along with a new sign-on system called Digits (launching in 218 countries and 28 languages today). While tweets will remain Twitter’s foundation, this is a real strategy shift that’s in many ways similar to Google’s growth out of search.
“It’s not a departure so much as moving beyond Twitter the product and moving into Twitter the company and the platform,” Twitter CEO Dick Costolo tells WIRED. “It’s about helping define the future of the mobile landscape and building an application developer’s platform for the future.”
While tweets will remain Twitter’s foundation, this is a real strategy shift that’s in many ways similar to Google’s growth out of search.
Here’s the thing. Many software development kits are, well, lousy. They often have pages and pages of instructions just to get set up. Fabric is dead simple by comparison. It’s just an extension that you add to your iOS or Android development environment toolbar. Click a button and a dropdown appears with buttons for its three different kits—one for adding Twitter functions, another for analytics, and a third for adding ads to your application. You select which features you’d like to include, Fabric scans your existing code and shows you where to place it. In most cases this is just a few lines of code. For Android, it can even insert itself with the push of a button. If you’re writing for iOS, you just copy and paste.
Today, Fabric is built from three kits with a total of seven tools that can work together or separately. The Crashlytics kit will offer tools for tracking crashes, analyzing app usage, and beta distribution—that is, a way to distribute apps to iOS and Android beta testers before launching them in the app stores. The Twitter kit will let developers add embedded tweets to their apps, offer a sign-in via Twitter function, and most interestingly, a new way to sign up for any app with a phone number instead of an email address. The last kit, MoPub, lets anyone add mobile ads to an app.
It’s a very, very different Twitter.
courtesy Twitter
Before Ashton Kutcher and Barack Obama and Justin Bieber took Twitter mainstream, you could just kind of wander into the company’s headquarters unannounced. You might pass a pile of dirty dishes in a mini-kitchen (really, not much more than a sink) when you walked into its small office in San Francisco’s South Park. You were often met at the door by the CEO, who would show you the entirety of the space with a shrug and a wave of his arm. There was no Twitter mobile app back then. Instead, if you wanted to tweet, you used SMS. The company has been relentlessly changing ever since.
Today, when you arrive at Twitter, you have to jump through a series of even-by-tech-industry-secrecy-standards cumbersome hoops before you can go inside. Security guards in the lobby demand to see your driver’s license. The elevators don’t have any buttons; the guards control the elevators remotely. On Twitter’s main floor in a gargantuan building in San Francisco’s Mid-Market district, you have to register (again) with reception, where they give you a badge that will, with an escort, let you in the door. Once again you pass by a kitchen, but this one is elaborate and gourmet and has a steady stream of chefs who ply the counters with fresh food. Two never-ending bowls of bacon beckon you from either side of the room. The sinks are in the bathroom where they belong. Dirty dishes are, apparently, illegal.
But all of it, all of this towering accomplishment, was built on a core of SMS. And mobile SMS has very much remained at the company’s center. As Twitter expanded globally, that turned out to be an advantage. It put it in a great position to sign up users on feature phones (and even dumb phones), and it meant that over the years, it has been able to build out lots of infrastructure to support SMS, and cut deals with mobile operators all over the world.
Today, when you arrive at Twitter, you have to jump through a series of even-by-tech-industry-secrecy-standards cumbersome hoops before you can go inside.
That in turn is letting Twitter roll out Digits, a way for developers to allow users to sign up for new accounts using just their phone numbers. “Digits’ capability is fantastic and is going to be something that surprises the industry,” Costolo says.
In much of the world, new users don’t have email addresses, which is why Twitter, Facebook, and many other apps use phone numbers for registration and authentication overseas. If you’re a smaller companies, say, one with a single app and two programmers in a garage, you want in on this vast market. And, in theory, signing users in with a mobile phone numbers offers a way to get your app to nearly everyone on Earth.
In reality, pulling off mobile sign-in has been super complicated and expensive for small developers—even here in the United States. You have to send out millions of SMS codes, reliably, at global scale to do it right. Given that SMS can cost up to six cents per message, that’s something many just can’t afford. So Twitter is offloading that whole process onto its own system, giving developers the ability to add in-app SMS registration by just adding a few lines of code, and its eating all the costs.
In other words, it has taken a remarkably expensive and complicated challenge and made it easy and free. That’s the entire experience of Fabric in a nutshell. You don’t have to read through pages and pages of documentation. There aren’t reams of development kit rules to follow. It just plugs in and works.
Let’s say you’re in Android Studio. You open the Fabric plugin and you’re greeted with a list of the various apps you may be working on—Twitter created the SDK with independent contractors and internal departments in mind so that you can keep apps from one client or team segregated from others.
You don’t even necessarily have to know how to write new code. It does it for you.
You can install the Twitter kit to an app with just a single click, and then select which of its options you want to include—Digits, Sign in with Twitter, or embedded tweets. Once you select Digits, Fabric will automatically insert everything you need to get it running on the sign up screen of your app’s code. (On iOS, it’s not an automated insert, but it does let you copy the code with a single click.) Once you’ve verified your changes, you’re good to go, and you can see everything you’ve implemented and how well it’s working (or not) in a web-based dashboard.
From the app user’s perspective, it’s similarly easy. Choose Sign up with Phone, enter your country code and your phone number. You’ll then get a text message with an authentication code, which you supply to the app. Boom. You’re registered. The developer gets the user’s verified phone number and an authentication token. (Twitter says it does not, contrary to some reports, hold the number to reserve a “shadow” Twitter account.)
If a developer wants to go back later and add more options, they just select those from the Twitter menu, and everything is added in. You don’t even necessarily have to know how to write new code. It does it for you.
“We show you the lines of code and where to put them,” says Jeff Seibert, the Crashlytics co-founder who now runs mobile platforms for Twitter.
Screenshot: Twitter
Apps crash. It’s what they do. And for developers, diagnosing why can be hugely problematic. Often, developers don’t even realize there is a problem or realize its scale. You just get crash reports, like the one from iOS, with often impenetrable numbers across the bottom. Or, worse, you don’t even know about crashes.
Crashlytics essentially figured out how to run at a very low level in your phone’s operating system. That way, when one of your apps crashes, Crashlytics wakes up, records those impenetrable numbers as quickly as possible, and then sends them to the developer when the user restarts the app.
It’s a powerful tool that has been widely embraced. (Crashlytics is running in apps used by more than a billion people every day.) Developers, like Strava, sing its praises in public. What Fabric adds to that are tools called Answers and the new beta distribution capability. Both of these were rolled out earlier this year, but now all three are bundled in Fabric.
The beta distribution option will let developers recruit testers from any platform, just by sending an email. Crashlytics helps developers diagnose the severity and scale of their application crashes. It ranks them by severity (typically, the number of people affected) and can even diagnose hard to suss out information—like whether or not the user is running a jailbroken iPhone. Answers is like Google Analytics for your app. It can tell you how many daily and monthly active users you have, how many new users you’re getting every day, the number and length of user sessions, and how popular your various versions are and how many people are using each of them.
One source familiar with Twitter’s strategy compares the move to Google branching out from search into YouTube and Android and Chrome.
And finally, there’s the MoPub advertising kit. It lets anyone monetize an app by joining its exchange and dropping a few lines of code. You can then opt for takeover ads (but please don’t!), banner ads, or “native” ads that match the layout and format of your application.
And if Crashlytics was the impetus that gave Fabric its start, MoPub helps explain why Twitter would want to transform its business. If it can get people to use Fabric to run Crashlytics, it’s just a click away from having them also use MoPub—and giving it a cut of that advertising revenue.
Yet it also raises the question of why Twitter feels the need to branch out like this. It could be seen as a sign that the core business isn’t growing quickly enough. Twitter, of course, rejects that interpretation.
“We’re growing approximately 100 percent year on year, Weil says. “We’ve built a strong consumer product that touches every corner of the Earth. We’ve built on top of that an ad business that’s growing very quickly and shows no sign of slowing down.” That very growth is fueling this latest expansion, he says.
One source familiar with Twitter’s strategy compares the move to Google branching out from search into YouTube and Android and Chrome. “You could ask what does Gmail have to do with search? How does it help search? And the answer is it doesn’t. Gmail adds value to Google, though, because it makes Google more central to people’s lives.”
Twitter’s hope is that all of these features will mean that Fabric will be where developers go first when they start building mobile apps. That it will be a sort of home base. “You can think about the beginning of the web landscape similarly,” Costolo says. “There were all these sets of different services you had to build into your app, or try to roll your own,” Costolo says. “As the ecosystem matured, they became a robust simple service layer.”
That’s what Twitter is trying to do. It wants to be the simple service layer that lives at the bottom of every application you use.
“Right now Twitter the product and Twitter the company are synonymous,” Weil says. “After this, Twitter will now be Twitter the consumer app, and Twitter the mobile services company.”
It’s been a long road. And it still has much to do to win developers’ hearts. But the kitchen is huge now, and there’s plenty of fat on that bacon.
No comments:
Post a Comment