My analysis of the game behind Google Pixel and Google Home. Kinda long read! Cross posted from my blog
The Age of Speech Recognition, Finally!
People have been working on perfecting speech recognition since the last 60 years. At first, the technology relied on recognizing sound patterns and started with identifying digits. Then the use of efficient algorithms made it possible to understand a few hundred words. The next breakthrough came with the use of Machine Learning techniques like HMM’s which would efficiently guess the words they didn’t recognize, based on assigned probabilities learned from training samples. In spite of significant technical advancements, speech recognition suffered from two major deterrents - the lack of humongous training data and a real pain-killer use case. Things changed rapidly over the course of last eight years with the advent of the smartphone because typing became a real problem which needed to be solved. Secondly, Google had been building an enormous repository of training data based on search queries and text from web pages and was now ready to deploy its voice recognition capabilities to solve an omnipresent need.
A New Way to Interact
Typing is a painful process. Firstly, you have to learn to use a keyboard. Have you ever seen your parents use the keyboard, typing one character at a time? It’s painful to watch. Secondly, typing is not as natural touch, scroll and pinch are. (You must see the video of how a two-year-old interacts with the iPad). Speaking, on the other hand, seems a more natural way of instructing as this is the way we communicate with people around us.
With the advances in speech recognition technologies, not only did speech start replacing typing, it opened up a plethora of possibilities - You could now interact with your devices hands-free. It began with an existing use case of when your phone is in your pocket, and you are in a situation where pulling it out is not a possibility, for example, while running or when traveling in a crowded subway. You could just tell your phone to do something. However, the brilliant people at Amazon thought of the world where you will have devices which are always listening and would provide you assistance with any information you need. It felt like asking for help from real human and thus was born the era of having virtual assistants.
A Paradigm Shift
Since the last 20 years, we have been searching for information. Google shows us ten results with a few ads. It’s left up to us to decide which of the results answer our query. Sometimes we click on ads because they seem highly relevant to our question. It’s a win-win situation for both.
But the next decade is going to be orthogonal. We will ask our assistants for information, but we will expect the assistant to give us the correct answer, not ten options. This shift is significant for Google which earns 10’s of billions of dollars from showing ads. If they don’t get onto this bandwagon, it would be like the Kodak moment where they totally missed the bus on digital cameras. Not only Google but everyone else, including Apple and Amazon, want a piece of this pie. Apple already has Siri on every iPhone and (now) Mac, and Amazon is launching smaller echo devices for every room.
Who owns the touch points?
It’s always been about owning touch points. Google was the default search engine in Mozilla. But it realized the potential threat of being ousted as the default and hedged it by launching one of the best browsers out there. When you own the browser, you own the touch points to information access. The same was repeated in the era of smartphones where Google is the default search engine on an Android phone. However, the touch points are now becoming more and more distributed. They are not just limited to your phones or laptops but have moved onto your living rooms as stand-alone devices. Thus it makes sense for Google to go after the in-home-virtual-assistant market with Google Home.
On mobile, Apple already has Siri and is adding many software services like iCloud and Apple Music because they anticipate the next wave of growth coming from subscription based services. Samsung, on the other hand, has a war chest of billions of dollars, built from selling high-end Android devices, and there is no one stopping them from creating their own virtual assistant and cut Google out. Google can’t afford this, and there are two ways to circumvent the problem - either own the phone hardware business or do what you did with Android, provide the OEM’s with a smart virtual assistant to use out of the box. In my opinion, Google Pixel phone is an attempt at both - own a pie of the hardware business and establish the use case for high-end things virtual assistants and VR where a tight integration with hardware is a required. Once the use cases are established, every other manufacturer will want these technologies, which Google will happily bundle with Android, given you have a compliant hardware.
War of Ecosystems
If there is anything noticeable happening, it’s a fact that Apple which is known for making hardware products with supporting software, if now foraying into software services like Siri, iCloud storage, and Apple Music. While on the other hand, companies like Google and Amazon, which are known for their software services are getting into hardware with products like Echo and Pixel phones. All of this is being done to lock you in an ecosystem where the value of continuing will far exceed benefits of making a switch. For example Amazon Prime has a music service, video streaming bundled with next day shipping, while Apple’s suite of services offers storage and music. Google’s ecosystem will provide the intelligent assistant across phones and home devices and will include free storage of photos and maybe Google Music. All of this is moving towards a subscription model which will provide you everything - storage, music, movies and lot more. The sooner a company can get you into their ecosystem and create enough value to act as a barrier to switching, the more value they will create in the long term. So don’t be surprised if any of these companies go on to buy Netflix or Spotify.
Who will be the winner?
When we talk about networking, it’s a winner takes all type of a game. However, with regards to ecosystems, there can be mutually sustaining ones just like Android and iOS. Google has the advantage of an already built knowledge graph so it’s assistant would be much more useful than Siri or Alexa. However, it will still rely on Amazon to complete the transactional queries (real purchases involving real money) and will have to figure out a new business model in the ‘voice’ search world, where users will expect the correct answer and not a set of links. Amazon, on the other hand, has a unique advantage of being able to ship almost any product on earth to your doorstep and is quietly building an intelligent assistant and other services to keep you hooked. While Apple has the advantage of an iPhone being in the pocket of millions of people, it like Amazon, still has an assistant with limited capabilities and a new threat from Google in the high-end phone market.
Each player has a unique set of advantages and shortcomings they need to fix. What is exciting is the fact that even after being in business for almost two decades, none of them enjoy a monopoly. They will have to sweat it out if they want to stay relevant in the next ten years, which in turn implies innovative products at the lowest possible price for us consumers.