On September 19, Google announced it had acquired API.ai, a startup that launched Assistant in 2011, one of the earliest independent virtual digital assistants (VDAs). The acquisition has plenty of industry followers wondering why this occurred. Blog posts about the acquisition from both Google and API.ai are, in the Google tradition, vague about what will happen, though API.ai’s executives said in their blog posts that it will be business as usual for their customers.
The Evolution of API.ai’s Offerings
API.ai, originally known as Speaktoit, had some success with Assistant, which was downloaded more than 10 million times from the Google Play store. But in December 2014, the company pivoted to focus on becoming a developer platform for natural language processing (NLP). The platform enables developers to create their own voice apps across a range of end markets, including the mobile device, wearables, robotics, automotive, and smart home markets. API.ai licenses its technology in tiers based on the number of queries per month, from $0/month for up to 6,000 queries to $1,100/month for up to 2.4 million queries per year. Pricing for over 2.4 million queries per year is on a case by case basis. API.ai has been exploring a broad range of opportunities, including Slack bots, connected cars, and smart short messaging service (SMS) chatbots.
Google’s Latest VDA Strategy
Tractica’s report on Virtual Digital Assistants described API.ai’s strengths as follows:
- Platform-agnostic strategy
- Scalable automated business model
These are the kinds of strengths that resonate in “Google-land” for any type of business it thinks about, but does Google really need NLP intellectual property? Isn’t Google generally recognized as one of the world’s leaders in artificial intelligence? There must be some other driver for the acquisition. Some clues are found in an excerpt from Tractica’s assessment of Google’s VDA strategy in our report:
A lot can be revealed about Google’s VDA strategy from a recent company blog post from CEO Sundar Pichai on May 2: “Looking to the future, the next big step will be for the very concept of the ‘device’ to fade away. Over time, the computer itself – whatever its form factor – will be an intelligent assistant helping you through your day. We will move from mobile first to an AI-first world.” Google’s intent is for its VDA to become not only cloud based, but context aware as well. According to a Google blog post on May 18, “The assistant is an ambient experience that will work seamlessly across devices and contexts. So you can summon Google’s help no matter where you are or what the context. It builds on all our years of investment in deeply understanding users’ questions.”
But we are not there yet. Google Now, which will transform into Google Assistant, will exist and live in some separate spaces for now. At Google I/O in May 2016, Google announced that Google Assistant will be built into its Amazon Echo competitor called Home. It will also live in a new Google chat app called Allo. During the conference, Google announced several third-party partners, but did not discuss the possibility of a third-party application programming interface (API) platform.
Now, take a look at what Tractica said about Apple’s vision for Siri:
Since debuting in the iPhone, Siri has been added across a range of Apple products, including the iPad, iPod Touch, Apple Watch, and fourth-generation Apple TV…. At Apple’s Worldwide Developer Conference (WWDC) in June, Apple announced Siri will be introduced to Mac computers in its new OS update called Sierra in 4Q 2016, which means every Apple device moving forward and the vast majority of the installed base of Apple devices will eventually feature Siri.
… Siri will now have an opportunity to help consumers in every category of device Tractica has identified for consumer VDAs (with the exception of fitness trackers): smartphones, tablets, PCs, automobiles (Apple CarPlay), smart watches, and smart home (Apple TV or iPad, both of which will run Apple’s Internet of Things (IoT) app HomeKit).
… on June 13, 2016, Apple announced it will launch SiriKit, APIs that will open Siri to Apple’s vast network of third-party developers. This strategy of Siri-powered apps puts Apple and Siri in an enviable position. By leveraging the Apple developer community, Siri will be one of the primary sparkplugs for consumer VDA innovation because of the lack of friction/fragmentation in the Apple ecosystem.
Future Plans for Google Assistant
API.ai is Google’s entrée to embedding Assistant into third-party apps. The platform enables Google to extend Assistant across a broad range of apps, first in Google Play and Chrome, but far beyond that to iOS and other platforms, mobile devices, PCs, and device-agnostic devices. Google Assistant can potentially be white-labeled within other apps, perhaps showing a future not dominated by only a few big names in VDAs, but as the embedded secret sauce behind thousands. The platform also reinforces Google’s vision of eventually offering completely cloud-based, app-less assistants, but in this case, it is Google-branded and available to users however and wherever they want.