SwiftKey’s Timely Exit

Another European startup exits to an American tech giant. Predictive keyboard app maker SwiftKey, based in London, confirmed today it has sold to Microsoft. Sources put a price-tag of $250 million on the deal, although neither SwiftKey nor Microsoft are confirming a figure.

It’s interesting that such a widely used app — with installs on some 300 million devices, across Android, iOS and its SDK products — should feel the need to take an exit. Although it seems likely acquisition was always the end-game for SwiftKey, given that predictive keyboard tech is (arguably) an expected necessity rather than a nice-to-have add-on. And given that a keyboard app has no obvious path to becoming a platform in its own right, as certain messaging apps have been able to.

Basic predictive text has been around since the pre-smartphone era. And while SwiftKey has always trumpeted claims of near telepathic powers of prediction thanks to its AI/machine learning core tech, it’s arguably been a shrinking differentiator with mobile OS platforms upping their own game in that regard.

For example, Apple added its own next word prediction feature to iOS 8 back in June 2014 — even as it also opened up the iOS keyboard system wide keyboard to third party devs like SwiftKey for the first time. The latter move substantially increased the number of keyboard competitors piling onto SwiftKey’s territory. And while SwiftKey has grown active usage from some 200 million+ devices as of mid-2014 to some 300 million now, a year and a half later, the opening up of the iOS walled garden would appear not to have led to the influx of usage that SwiftKey might have hoped.

Asked by TechCrunch in mid-2014 whether it was gunning to become a billion-user business, co-founder Ben Medlock suggested that sort of scale was not “completely out of range”, given what he said were roughly 300 to 400 million users on iOS that would soon be coming within the orbit of SwiftKey’s app.

“If you look at the proliferation of touchscreen devices I think there really is a chance of getting to a billion end-points,” he said at the time. “Using the technology, we currently have somewhere north of 200 million [active devices] — that’s through a combination of our licensing deals and our app downloads… And we’re just ramping up now to really focus on distribution.”

In the event, SwiftKey has thus far managed to achieve around third of a billion users. Most of whom, presumably, are on Android. (Although it does not break out usage per platform SwiftKey has been on Android since 2010, and iOS only since fall 2014.)

Over some eight years in existence as a business, the London-based startup raised just under $22 million from investors which include Index, Accel and Octopus. So paying back those VCs with a timely exit was evidently the clearer path here, vs attempting to sustain and grow a standalone app business in an increasingly competitive landscape. Or trying to transform its app into a more valuable platform play.

In previous years SwiftKey charged for its app, but as keyboard competition proliferated it switched to a freemium model, in mid 2014 — offering the core tech for free and launching a themes store as its initial content monetization plan. But selling colored keyboard wrappers never looked like a hugely serious business model for an app with hundreds of millions of productivity-loving users. Even as giving away its core tech undermined the future viability of its existing licensing business to OEMs. (SwiftKey couched this as a way to widen its reach in emerging markets where its licensing model would have been out of reach anyway.)

And while it claimed to have other paid for content plans up its sleeve, its focus seemed more on acquiring new users by giving its app away than on milking its user-base. Which makes sense if acquisition is your end game because then you are playing a pure numbers games — since you’re selling potential reach to your buyer (vs your rivals’ potential reach).

Of course the basic math here, from Microsoft’s side, is calculating how far it might be able to spread its Cortana AI assistant if it’s piggybacking on the existing SwiftKey user base vs how much it’s having to spend to buy SwiftKey’s reach. But the fact Microsoft’s interest in SwiftKey apparently centers on Cortana does beg an interesting question of why SwiftKey didn’t move to develop an AI assistant of its own — to augment and extend its predictive keyboard app.

And perhaps transform into that sought for platform. Since there are clearly some synergies to be leveraged between predictive keyboard technology and a virtual assistant designed to crunch data to make useful predictions, as Microsoft has recognized.

Why not do that? In a word: risk. When faced with the choice between the simplicity of an exit to an existing tech giant or the complexity of developing, marketing and maintaining another product that’s at least one step removed from your core expertise — and which also brings you into direct competition with multiple tech giants already battling to dominate this space (e.g. Apple’s Siri, Facebook’s M, Amazon’s Alexa, Google Now, Microsoft’s Cortana…) — well, it’s easy to see why taking the money was by far the simpler choice for SwiftKey’s co-founders.

SwiftKey the product, then, was likely a very well-honed, VC-backed feature looking to sell itself to an established tech business. As today it confirmed it has. This exit suggests that sustaining a standalone app business is not easy in today’s crowded app marketplaces — and that even with hundreds of millions of users monetization can still pose challenges. But it also underlines that the right strategic choices can pay off pretty handsomely.

As SwiftKey’s co-founders put it in a rather understated statement on the acquisition: “We believe joining Microsoft is the right next stage in our journey.” In other words, if your exit strategy is, well, to make an exit, that’s not necessarily a commentary on how sustainable or otherwise your standalone app business might have been. Although having a pay-back strategy is a must. So perhaps it is a commentary on the relative ease of an exit vs pivoting in the hopes of future value.

A ~10x exit isn’t too shabby at all. Plus it’s guaranteed money in the bank vs the unquantifiable risk of spending to keep swimming in the seethingly competitive cauldron of the smartphone space.

So not quite a swift exit for SwiftKey then, but a timely one that will have its investors smiling nonetheless.