Voice Assistants Alexa, Bixby, Google Assistant and Siri Rely on Wikipedia and Yelp to Answer Many Common Questions about Brands
Consumers are increasing employing search through voice assistants and those queries extend to questions about brands. In the Voice Assistant SEO Report for Brands published this week by Voicebot in collaboration with Magic + Co., a key finding of note is the particular importance off Wikipedia for voice search results. Voicebot researchers asked over 4,000 questions to Alexa, Bixby, Google Assistant, and Siri about 200 consumer brands and one of those formulations was, “What is [BRAND NAME]?”
For 88% of the correctly answered brand name queries, Samsung’s Bixby consulted Wikipedia. Alexa, Google Assistant, and Siri all can thank Wikipedia for 99% of their correct answers to these questions. There is a great deal of industry discussion about optimizing your voice app to answer questions using Nameless Invocations for Alexa and Implicit Invocations for Google Assistant. Google’s vaunted knowledge graph and answer boxes are also cited often as the key to voice search engine optimization. However, for brands, Wikipedia is the information source most referenced to answer the most basic question about who they are.
Voice Assistants are Optimizing Not to Fail
There are many information sources that could answer the “What is [BRAND NAME]?” question including digital properties owned by the brands themselves. Every one of those brands has a website and Voicebot researchers even found that 26.5% of the brands in our study had an Alexa skill and 21.5% a Google Action. However, brand websites do not have a common structure and using automation to find a simple answer to a fact-based question is problematic.
Wikipedia, by contrast, is structured consistently and its volunteer editors typically offer a simple description of a brand or other topic in the first sentence on the page. This allows voice assistants to consult a single source and offer their overseers high confidence that users will almost always receive a correct response. In addition, the voice assistants can also lean on their text-to-speech capabilities to offer to read more information after the first sentence is complete. Company websites might have images, charts, headlines, or other content that is well suited to visual consumption but a poor user experience for audio. However, just because Wikipedia will not produce a technical error in answering the question, doesn’t mean it will optimally fulfill a user’s intent.
Voice assistants are using Wikipedia today because they are optimizing not to fail instead of optimizing for success when questions are posed about brands. You can see this in the data. Google Assistant far outpaces its rival voice assistants in providing correct answers for three of the four question categories. The difference is the smallest for the “What is [BRAND NAME]?” question. Whereas Amazon Alexa trails Google Assistant by 50-60 points for all four questions about brands, the shortfall is only 15-16 points for the question where both rely on Wikipedia as the primary information source. Using Wikipedia minimizes the difference in correct answers across voice assistants even if it doesn’t always deliver the best answer to a query.
Yelp is a big Help for Alexa and Siri
Alexa and Siri trail Google Assistant’s prowess by a much larger deficit when it comes to the local query, “Where can I buy [BRAND]?” However, when Alexa and Siri are able to offer correct answers, it is most often courtesy of Yelp. It is worth noting that both Alexa and Siri fail more than they succeed when asked these queries that typically rely on local search results. However, when they do succeed, Yelp is the source 89% of the time for Alexa and 81% for Siri.
Yelp offers a single source of structured data containing information about business locations similar to Wikipedia’s role for general information questions. The consistency of the structured data and general reliability of these information sources makes them more appealing than a company or other third-party website that have far more limited information or inconsistent data representation. And, the voice apps launched by many of these brands don’t necessarily provide the information required to answer the question. If the information is in the voice apps, it may not be clear to the voice assistant that it is available in a format that will successfully fulfill the user need.
Voice SEO Insights for Brands
Ben Fisher, CEO of Magic + Co., points out that voice apps have certain objectives that may not always answer consumer questions and brands should recognize that voice assistants will tap into multiple sources of information:
A properly designed Voice App can take control of the conversation for certain things, other things consumers can best be served by other sources. At the end of the day, it is about getting the consumer the answers and help they need, not just directing traffic to your app.
The research results suggest practical steps that brands can take to boost their voice search success is to shore up their Wikipedia and Yelp profiles to ensure basic questions can be answered when consumers ask. They also can incorporate some of this information into their voice apps to provide voice assistants a truly conversational source for answering questions about their brand.
Follow @bretkinsella Follow @voicebotai
New Data on Voice Assistant SEO is a Wake-up Call for Brands
Voice Assistant Timeline: Chronicling a Revolution from 1961-2019