How Amazon Alexa and ‘Voice Apps’ Are Changing (Local) Search
June 29, 2016 | Contributed by: Greg Sterling
Perhaps more than Siri, Amazon’s Echo/Alexa is popularizing and accelerating adoption of virtual assistants and voice search. And while the rise of voice search and virtual assistants doesn’t yet hold radical implications for marketers, traditional search and mobile consumer behavior is evolving as a result.
Roughly a year ago Amazon opened its Echo/Alexa assistant to third party developers. Third party services on Alexa are called “skills” (too bad it isn’t spelled “skillz”). Effectively, skills are “voice apps.” Earlier this month Apple opened up Siri and Maps to third party developers as well.
Earlier today Amazon announced that “tens of thousands of developers are building skills for Alexa, and there are over 1,400 new skills for Alexa—including Lyft and Honeywell, which were added today.” So far there are only a limited number of transactional capabilities. For example, Echo/Alexa owners can access Kayak travel information but can’t book reservations — yet.
Users can request a ride on Uber or Lyft, order flowers from 1-800-Flowers or a pizza from Dominos. They can also conduct local searches via Yelp on Alexa (category searches yield four listings typically). And there is a growing list of other things users can do with Alexa/Echo, including controlling smart home devices.
Amazon Echo has become a surprise hit and is now outselling Amazon’s Kindle devices. Amazon CEO Jeff Bezos told investors last quarter that “Amazon devices are the top selling products on Amazon … Echo too is off to an incredible start, and we can’t yet manage to keep it in stock despite all efforts.”
As indicated, there are more than 1,000 “skills” or voice apps. (What we might call skills discovery is already a problem.) The following is a list of “fun facts” that Amazon published to indicate developer and usage momentum:
- Since January 2016, selection of Alexa smart home API skills has grown by more than 5x
- Amazon now has over 1,400 Alexa skills—growing the catalog by 50% in just over one month
- Customers have made over 3 million requests using the top 10 most popular Alexa skills
- There are now over 10,000 registered developers using the Alexa Voice Service to integrate Alexa into their products
- There are tens of thousands of 3rd party developers currently working on Alexa projects
- Some of the most popular Alexa skills are Jeopardy!, Daily Affirmation, Magic 8 Ball, Fitbit, and The Bartender
The success of Echo prompted Google to quickly build and announce its Home device, complete with a budding developer ecosystem. Apple too is reportedly working on a stand alone assistant device. Microsoft is also likely to build one. In fact, Google is replacing the one-way notion of search with a conversational two-way “assistant” metaphor.
Echo has clearly captured people’s imaginations, as Siri did initially. Collectively, Google Now/Asssistant, Cortana, Siri and Alexa are training people to conduct voice searches. Microsoft has said that 40% of voice searches carry local intent. Voice queries are growing significantly on mobile devices and will ultimately overtake conventional text input search queries, just as mobile searches have overtaken desktop query volumes.
The move to voice search/virtual assistants does two primary things. It requires publishers and marketers to start thinking about content in terms of questions and how to optimize for voice search and rich answers. It also accelerates the need to integrate booking and transactional capabilities more quickly and broadly. In mobile and voice search contexts people are seeking to complete tasks rather than looking for lists of links. In mobile search, discovery and transactions go together because of the immediacy of intent in many mobile search scenarios.
Voice search and virtual assistants don’t work well for all search use cases. You don’t want to hear voice playback in a high consideration scenario such as vacation planning or car buying. There you need a screen to consumer content and evaluate results (which the Alexa app provides). Yet the growing sophistication of machine learning/AI and natural language processing make voice search an increasingly viable alternative to traditional search results if precise or rich answers can be provided.
Amazon is already beating Google when it comes to product search. The company could make further inroads as Echo/Alexa becomes a more capable general and local search alternative, especially as discovery becomes more sophisticated and transactional capabilities (e.g., restaurant reservations, travel booking, etc.) are enabled.
It’s clear that we’re now in the midst of another technology transition. Part of that involves location, context awareness and machine learning. But it also involves the transition from traditional one-way search and discovery tools to assistant based “conversational” and transactional search.