Amazon, Apple, Google, Microsoft and a host of startups are battling it out for dominance of voice as the next UX and for creating a truly intelligent personal assistant. Each has taken a difference approach to voice, which will determine the strength of the network effects at play and ultimately, the success of their products and platforms.
Apple, ever the high water mark for quality control, has created a completely closed service with Siri. While this makes interacting with Siri simple, it also severely limits its functionally. For now, it is essentially a search tool with fancy formatting for a few common requests, like asking about the weather. While they may slowly open the service to 3rd party developers, Apple has traditionally made money from selling applications – it is not clear that this will be a tenable model with voice.
Google has initially shown less interest in voice as a user interface, but given their strength in Natural Language Processing (NLP) they could easily be a major player. Their acquisition of Nest shows a move into the home automation market, which has been the earliest use case for voice. If they sell the hardware to the home, they could justify voice expense with hardware sales and data sharing agreements with 3rd parties (eg. Utilities). However, if they open the platform, it is more likely they get pushed into their core business model of advertising. Adverts playing through a speaker in the home will be a tough sell and may challenge Google’s ability to compete in the space.
Amazon meanwhile, has created a platform called Alexa with two SDKs to allow development of voice applications (Alexa Skills Kit) and integration with 3rd party devices (Alexa Voice Services). While there is no immediate monetization mechanism for developers, Amazon has created several incentives to build applications that should defeat the chicken-and-egg problem. First, they have given free compute space on AWS to developers so the variable costs for developers are very low. They are uniquely well positioned to offer this incentive given the scale economies already achieved in their cloud compute environment. And second, they have dedicated up to $100m dollars in seed funding for companies building applications for Alexa.
The indirect network effects for Alexa are strong and fundamental to value creation. At $200, the Amazon Echo (the primary device supporting Alexa) is an expensive joke machine and weather forecaster. Soon, however, applications like yelp, OpenTable and Uber could bring Alexa much closer to the long sought after personal assistant. Eventually, direct network effects could be built into the platform by enabling communication and scheduling via Alexa (eg. “Alexa, ask my sister if she is free for dinner tonight?”). Additionally, multi-homing costs are high due to expensive hardware and integration with other devices in the home. This competitive moat should make any platform that gets traction much more valuable.
Alexa creates value for users and developers. However, Amazon does not make money on the device or app sales. Amazon is uniquely well suited to build a valuable platform in this space because it never has to make money from the platform itself. Instead, value capture for Alexa is in Amazon’s existing businesses, AWS and Amazon.com. Every time someone uses an application service on Alexa – the code supporting that application runs on AWS, for which Amazon charges. Additionally, ordering items via Alexa is seamlessly integrated – driving volume in the retail business. Finally, integration with FireTV and kindle will allow for better control and consumption of content on Amazon devices – where Amazon can then capture additional value.
Amazon will win the voice platform contest because it is not restricted by the draconian governance of iOS nor is it tied to Google’s value capture via advertising. Once the flywheel of indirect network effects gets going – developer and user adoption will boom. Amazon will deliver more and more value through Alexa, while capturing value elsewhere.