Alexa and the conversational interface
Amazon is driving the smart home forward.
Amazon’s Alexa is winning the smart home because it does what other consumer-focused internet-of-things (IoT) products don’t: Alexa provides real value. Many companies have declared that they’re making an IoT platform, a standard upon which all IoT products would operate. The scale and frequency of such declarations show both the value of a platform and the current inability of any single company to successfully create one. Most companies are trying to conquer the smart home market with brute force: by getting the most devices in the most homes as quickly as possible and then telling the IoT market to follow suit. Moreover, they’re dividing their efforts between the smart home and the broadest possible IoT solution.
With Alexa, Amazon brings an elegant, focused solution: the conversational voice interface for the smart home.
“In our home environments, we don’t need more disruption. We don’t need more screens. If it’s any more complex than [the conventional light switch], it’s probably not going to get wide adoption.” — Yves Béhar
The value of voice
Voice solves a critical challenge for connected products in the increasingly-fragmented smart home: the user interface (UI). A difficult challenge for IoT devices is how to increase device functionality without the UI taking up valuable space or ruining a beautiful design or being overly simplistic to the point of confusing the user. After all, how many additional buttons and knobs and lights and displays will a user tolerate from a device that’s supposed to make their life easier and their home more attractive?
So if the UI can’t be built into the product, what’s the alternative? Since smartphones are ubiquitous, it’s natural to design smart home products with “an app for that”. However, apps make the simplest IoT devices far more complicated. My favorite example of awful user experience (UX) for smart home devices is the connected light switch, which requires the user to (1) unlock the phone, (2) open the app, and (3) turn on the light — far more cumbersome than walking across the room to turn on a conventional light switch.
So if it’s problematic using clunky UI’s and mobile applications to control the smart home, then there’s good reason to believe that the smart home needs a voice interface.
We can follow the progression of UI for previous technologies to see why voice applies so well to IoT. The original interface for computers was the keyboard, a tedious experience that required the user to carefully type precise commands. Apple introduced the mouse as a simpler UI to let users point and click, which is still the primary interface for laptops today.
The first BlackBerry smartphones had cumbersome built-in buttons and keyboards. Again, Apple simplified the smartphone UI with the touch screen, allowing each app to design a custom interface. An app is a great way to manage the intricate details of an IoT device, like setup and administration, but apps are painful for our simple, regular day-to-day interactions with home IoT devices.
The progression of UI has been to get increasingly lighter and out-of-the-way. Voice continues this trend and is a compelling solution for consumer IoT because of its simplicity: voice is hands-free, instantaneous and easily customized to each device.
Of course, a voice interface is only valuable if it engages the user, and there’s good reason to believe that voice has only reached a tipping point in quality just recently. Many factors are driving this, notably rapid improvements in natural language processing, the ubiquity of microphones, and the power of machine learning algorithms. As Tim Tuttle, CEO at MindMeld, tweeted about automatic speech recognition (ASR), “Accuracy improvements in the past 2 yrs have dwarfed all improvements over the past 30 yrs combined”. Moreover, Amazon CEO Jeff Bezos defined 1 second of latency as the key metric for Alexa, which underpins the ease and comfort of the experience.
Regardless of the underlying technology, what really matters to consumers is that now voice just works.
I was unable to find a detailed study of the quality of Alexa’s natural language processing, so I can only go by my own experience with its speech-to-text and interpretation of my commands. Almost all of the time Alexa is able to understand the words I say, even if it can’t always execute the commands I want. But Alexa is learning. Every week I receive an update from Amazon announcing enhancements to Alexa’s abilities to execute my commands as well as integrations with other smart home devices.
Smart home devices will struggle without a voice UI
Those of us in the IoT community are constantly seeking killer use cases, and it’s impossible to know what will work until it hits the market. However, I’m willing to bet that Voice UI solves a critical part of the IoT puzzle.
So regardless of any other features or capabilities, if you’re designing consumer products for the smart home or making bets on which IoT products will win, make sure that your device has a voice UI — or risk being yet another flop in the consumer IoT marketplace.
Thanks for reading! :) If you enjoyed it, hit that heart button below. It would mean a lot to me and it helps other people see the story.
This post originally appeared in the blog for Orange Silicon Valley