Software developers attend Twitter’s Flight conference.
In December of 2012, four days before Christmas, Twitter flew Jeff Seibert and Wayne Chang from Boston to the company's headquarters in downtown San Francisco. A year earlier, the two men had founded a startup called Crashlytics. The startup's namesake product is a bug-reporting tool, designed to help mobile software developers figure out when and why their applications crash.
The men had a hunch they were about to make a lot of money.
"We were just saying, if we get an offer, make sure you don't show any emotion," Mr. Chang said. "If it's a penny, if it's a trillion dollars ... always sound like you're a bit disappointed."
Their hunch proved correct. On Christmas day, Crashlytics became the most expensive acquisition in Twitter's history (the exact terms of the deal have not been made public, but the price tag is believed to be north of $100-million U.S.).
And yet, something about the deal didn't make sense. Twitter's core product is the world-famous microblogging service that today is used by a quarter billion people. It has little to do with a bug-reporting service for mobile developers. Even the Crashlytics founders were skeptical as to why Twitter was interested in their startup.
It wasn't until a couple of weeks ago that Twitter finally made clear the overarching strategy behind the Crashlytics acquisition. At a developers' conference in San Francisco, the company unveiled an audacious new product called Fabric. It is, essentially, a kind of Swiss Army Knife for mobile software developers and includes tools that do everything from crash reporting to ad-based monetization to user login streamlining. In other words, it provides turnkey solutions for many of the most cumbersome parts of building mobile software.
"This is about something much broader than Twitter the consumer application," says Kevin Weil, Twitter's vice-president of revenue products. "This is about Twitter Inc. becoming a mobile services company."
Why? Twitter's core user growth is slowing and its advertising reach isn't expanding at the pace investors would like to see. If the company can woo developers into including its product in everything they build, from apps to websites, it can ultimately reach more users and win over more advertisers. Facilitating development directly also offers new lines of business for its ad service.
It's no easy feat. With the strategy, Twitter is aiming to join the tech industry's highest order, where Google, Amazon, Facebook and Apple dominate in the race for big data. Their businesses allow them to gather detailed information about trillions of interactions – from e-mails sent, to pages searched, to posts liked.
Over the past decade, those massive stores of data – and the myriad algorithms designed to tease out the information from the noise – have become the corporate world's most valuable commodity. Not only do they hold the potential to solve all manner of business problems, they also help turn advertising from an art to a science. The data that tech giants glean from their users' habits and interactions allows advertisers to pinpoint exactly which people are most likely to buy certain products.
In order to collect that data, the biggest tech companies have all positioned themselves as middlemen of one kind or another, building the platforms that facilitate everything from web search to e-commerce to social media. It is exactly that sort of company – a provider of a platform, not a product – that Twitter is now trying to become. And the customer base it has in its sights isn't everyday users: It's the people who build mobile software, who Twitter hopes will ultimately give the company access to everyone who uses the Internet.
After eight years as a social media player – and facing an increasingly frustrated investor base – Twitter is setting its sights on scaling up to be one of the world's biggest technology companies. But to do so, it must become something more than the name behind a 140-character blogging service. It must become a platform provider, a data giant.
Reaching advertisers
Twitter's world headquarters occupy a few floors' worth of office space in a downtown San Francisco office building complex. It is, by tech industry standards, a modest corporate campus – outfitted with a few employee lounges and an in-house cafeteria. The various food and drink stations are often labelled with hashtags (the sign atop the carvery reads: #meatzone). Posters and other visual mementos of tweets that became particularly popular over the years line the walls.
Although it has only a fraction of their market capitalization, Twitter does share key elements of its corporate strategy with its tech giant neighbours in Silicon Valley – chiefly, a focus on amassing users, even at the expense of short-term profit.
That focus partly explains why, almost as soon as it bought Crashlytics, Twitter started giving the product away. Prior to the acquisition, Crashlytics charged developers a fee for some of the bug-reporting functionality. But Twitter quickly began to offer the entire product at no cost – turning a profit was not an immediate priority. Two years later, the same strategy is visible in the Fabric toolkit (to which the Crashlytics software now belongs).
"We don't judge this by revenue," Mr. Weil says. "The way we judge success with Fabric is, are we solving developers' core needs?"
In many ways, Mr. Weil is echoing what has become a popular tech industry refrain. In Silicon Valley, talking about making money has become, oddly, a kind of faux pas. Executives at many of the world's biggest tech companies are much more inclined to frame their business strategy as focused exclusively on making the user happy – profits be damned. It is this phenomenon that explains, in part, why Amazon, a $139-billion company, has only very rarely turned a quarterly profit in its 20-year existence.
While talking about putting the user first at all costs makes for good public relations, in reality, the user is where the profit ultimately is – even if that user pays nothing.
Every Twitter user is a data geyser. Not only does the company collect all manner of information on its users' location, usage rates and behavioural habits (for example, what sorts of people they are likely to follow on Twitter), it also mines the billions of tweets those users post for keywords. Send a lot of tweets from Toronto, and you might start to see ads for Toronto-based businesses. Send a lot of Tweets about yoga, and you might see ads for yoga pants.
And precisely because it allows for this sort of targeting, data are especially useful to advertisers. Like many of its corporate peers, Twitter makes most of its revenue from advertising. Currently, Twitter allows advertisers to target specific audiences based on about eight different categories, including user gender and location. However, that's not nearly as exhaustive (nor is Twitter's user base as large) as what's available on Facebook, for example, where there are more than a billion users and advertisers can hone in on everything from education level to their past or intended purchases.
"Big data advertising has the potential to be magical," says Neil Bearse, associate director of marketing at Queen's School of Business. "It has the potential to make people say, 'I had no idea I needed that, and you showed it to me at the very right time.' But for the marketer who knows their target, Facebook still offers a far more precise product."
Broadly, the value of social media advertising is based on two things – the overall size of the audience, and the specificity with which advertisers can target any specific demographic. As such, one of Twitter's challenges is fairly straightforward – it needs more users.
Twitter chief executive officer Dick Costolo likes to describe the company's user base as a series of circles. The smallest one, made up of about 280 million people, is the monthly active user base. Beyond that is the demographic that consumes Twitter's content, but doesn't contribute any tweets (Mr. Costolo estimates this number at another 250 million to 500 million). Until the Fabric announcement last month, the largest circle in the set comprised everyone who sees tweets embedded in news stories or other media across the Web.
Now, with its new set of developer tools, Twitter can add one more potentially much larger circle – an entire ecosystem of apps that aren't made by Twitter, don't necessarily run the company's core microblogging service, but still use Twitter-developed code.
Such an ecosystem could prove lucrative because developers who use Twitter tools are also much more likely to use Twitter's advertising service to monetize their apps. MoPub is a Twitter-run ad exchange that not only connects developers and advertisers, but also gives developers powerful tools to track exactly how much money they're making from those ads. Suddenly, Twitter's value proposition for advertisers changes drastically.
"Advertising is obviously one of the big key drivers behind all of this," says Greg Gunn, vice-president of new product growth at Hootsuite, the Vancouver-based company whose namesake product lets users control their social media presence on 35 different services, including Twitter.
"The whole strategy is about going, hey, Twitter's got a really great app, but if we can help provide a network of apps that our advertisers can reach, that's a very large and potentially hugely profitable business goal for them."
Luring users
Building out a new circle of potential partners, customers and users is especially important for Twitter because the company's smallest circle – made up of users who actually tweet on a regular basis – isn't growing as fast as investors would like. The monthly average user growth rate at Twitter has declined or remained flat every quarter since the company went public a year ago.
For years, Twitter tried to keep user growth high by making its service a sort of global water cooler – a place for everyone to talk about the news and culture of the day. The company spent considerable time and effort building relationships with Hollywood studios, news organizations and celebrities. And for the most part, the strategy showed results.
"If you want to talk about your TV show, you do it on Twitter," says Alfred Hermida, an associate professor at the University of British Columbia School of Journalism and author of a new book on social media, Tell Everyone: Why We Share And Why It Matters. "If you want to talk about your favourite sports team, you do it on Twitter."
But Mr. Hermida notes that such successes do little to offset the central and growing problem Twitter now faces – the divergent preferences of its casual and hard-core users.
Twitter's user interface has always functioned as a stream of sorts – what you see is whatever is happening at the moment you check the site, and there's no easy way to browse content any way but chronologically.
Twitter's veteran users mostly love that setup, but new users can often find it confusing. The company has tried to make the initiation process a little simpler, in part by adding features such as automated recommendations for who to follow. But even the slightest hint of more sweeping changes – namely, altering the default news feed to display tweets from people that users don't follow, or in any manner other than chronologically – inevitably prompts howls of outrage from seasoned users. And it is paramount that Twitter keep those users happy, because while many new users tend to "lurk" without posting many tweets, it is the veteran users who create much of the content on Twitter.
Twitter's new-found focus on building developer tools is aimed largely at keeping the company's user growth rate high by making a Twitter account an indispensable digital tool – even if the user doesn't spend much time on the microblogging service.
An apt analogy is Facebook Connect, a service offered by the social network that lets users sign in with their Facebook credentials on non-Facebook sites, like Netflix, Groupon or The Washington Post. Connect has quickly become one of the most popular ways of moderating the comments sections on myriad news sites, among other things. Even though Connect has only a tangential relationship with the core Facebook service, its widespread use has made Facebook an integral part of the wider Internet experience.
Or consider Android, Google's free-to-use operating system for mobile devices. Even though Google makes almost no money from Android itself, the software's popularity means Google code is running on the majority of the world's smartphones – and that, in turn, makes it more likely that smartphone developers will use other Google services as well (including Google's ad exchange, from which the search giant makes huge sums of money).
In effect, this is what Twitter is trying to do with its new developer tools – become an indispensable part of the wider web experience, and in the process encourage developers to embed other Twitter services – including Twitter-run advertising tools – into their apps.
"If they can become part of the fabric of everything people do on the Internet ... that's more reason for people to sign up for a Twitter account," Mr. Hermida says. "At the same time, that allows them to collect a lot more data on what people do and why they do it – and that's data they can sell to advertisers."
More data also allow Twitter to develop a range of new ways to customize (and derive revenue from) its core microblogging service. For example, Mr. Hermida suggests, the company could begin offering geographically tailored services – if there's a sudden rush of tweets about a train delay at Toronto's Union station, and Twitter knows a user frequently logs in from that location, it could alert that user, whether they follow the people tweeting about that train delay or not. Or, when a user visits a mall, Twitter could begin pushing daily deal tweets from the accounts of stores located in that mall – taking advantage of a recently announced Twitter "Buy Button" that lets users make a purchase by clicking on a link embedded in a tweet.
But that's all easier said than done. For now, Twitter has neither the user base nor the granular advertising ability of many of its large competitors – and virtually all those large competitors also have much deeper pockets. Besides vast differences in market cap and profitability (Twitter posted adjusted earnings of just $7-million in its most recent quarter. Facebook, in contrast, posted adjusted profit of $1.1-billion, while Google posted $5.36-billion), Twitter only has about 60,000 advertising partners, compared with millions each for Facebook and Google. Twitter's ad load – or the percentage of ads to overall content, is about 1.3 per cent, whereas ad load at Google and Facebook is between 5 and 7 per cent.
Twitter does have one advantage, however. Of all the biggest technology names that make up the industry's top tier, none were built from the ground up to focus on mobile computing the way Twitter was.
Courting developers
With the company's share price down more than 37 per cent so far this year, Twitter has been the subject of growing shareholder frustration in recent months – most notably, a number of big-name investors have publicly questioned whether the company's senior management has a clear plan for user and revenue growth.
But ever since it was conceived in early 2006, Twitter has always been an uncertain entity.
The idea behind Twitter was first proposed by Jack Dorsey, who was at the time an undergraduate student in New York and is now, since Twitter went public, a billionaire co-founder.
The service started as a side project at a company called Odeo, whose employees were predominantly focused on the startup's main mission: podcasting tools. Designed initially for internal use by Odeo employees, Twitter was built as a "social utility" – a fancy way of saying it was a tool for posting short status updates. Within a year, Apple's iTunes would effectively render Odeo's podcasting business irrelevant. But Twitter, which had neither competitors nor even a well-defined business case, survived.
From the beginning, the central idea behind Twitter was to build a service that let users broadcast SMS messages from their phones to a wide group of people – but beyond that, there was no real mission statement. The only constant throughout the design process was that the service was intended to be used first and foremost on a phone, not a desktop.
"Twitter has been a mobile company since the beginning," Mr. Weil says. "The 140-character limit came from text messages."
The advantages of having focused on mobile devices from the very beginning are clear today – most notably, in many of Twitter's newest products, including its developer toolkit.
By far the most quickly adopted part of the Fabric toolkit so far has been something called Digits, a ready-to-use user sign-in service. Instead of building a sign-in screen from scratch, a mobile developer simply uses Digits to let users sign-in by entering a phone number. The service then confirms a user's identity by sending a text message to that number. Digits not only has the potential to save developers time and effort, it also saves users from having to remember countless different e-mail, user name and password combinations.
A number of big-name partners have already decided to use Digits. McDonald's, for example, is including the functionality in its daily deals app – not as a replacement of its existing account creation process, but as a sort of first step, designed to lure users into using the app and then eventually sidestepping Digits and signing up for a full McDonald's account.
"From our perspective, [Digits] is a really interesting way to bring new customers into our world in a way that's lightweight and really reduces friction," McDonald's chief digital officer Atif Rafiq said at a Twitter developer conference.
"What we're trying to do here, if you think about it, is ... ask users to make the commitment that they're ready to make to enter our world and grow that relationship over time."
In return for giving companies such as McDonald's an easy sign-in service such as Digits, Twitter gets its product integrated into apps that run on millions of smartphones, while simultaneously making it more likely that the developers who use Digits will go on to use other Twitter services.
One of the main reasons Twitter was able to build Digits is the company's extensive working relationships with telecom carriers around the world. Indeed, "mobile-first" is at the core of Twitter's sales pitch for Fabric. The company hopes to become a mobile software developer's best friend. It is, in some ways, a sharp turn for a company that once rewrote its terms of use to effectively kill many third-party Twitter clients. But it is also necessary if the company is to revive its user growth and expand its network to the size and scope of the giants with which it will soon compete.
"We want to build tools developers will love, because then they'll find it easy to integrate Twitter into their apps, and ultimately that will help benefit the Twitter network," says Mr. Seibert, one of the Crashlytics co-founders who now serves as Twitter's director of mobile platforms.
The success of Twitter's planned shift from social network proprietor to mobile-services company is far from certain. Many mobile developers have become accustomed to the tools already offered by Google and Apple – and it may take a long time to convince them to switch, or at least include, Twitter's new services in their development process.
In the meantime, Twitter still has to contend with an increasingly impatient investor base that has seen the stock price rise sharply in the first few months following its initial public offering and then drop right back to where it was a year ago.
Analysts are taking a wait-and-see approach.
In a note to clients following Twitter's recent investor day, JPMorgan analyst Doug Anmuth wrote that, even though "expectations were extremely low going in," the company did not shy away from addressing its user growth problem. At the meeting, senior Twitter executives laid out a number of new initiatives to try to make the service more alluring to new users, including a feature called While You Were Away, which summarizes important Tweets posted since the last time a user logged in.
"To be fair, Twitter said a lot of the right things, but now they have to execute," Mr. Anmuth wrote.
But if the transformation works, Twitter might finally find itself occupying one of the most lucrative positions in the tech industry – that of middleman between creators and consumers, amassing users, data and advertising revenue in the process.
"If we are a key part of the development life cycle for every app [developers] build, there are going to be a lot of opportunities that spring from that," Mr. Weil says.
"This is the first step."