Is voice search speaking up?
Figures are floating around affirming that voice search has proliferated people’s lives to a point of no going back. 65% of Google Home or Amazon echo owners “can’t imagine going back” says GeoMarketing.
25% of all Windows 10 desktop searches are done via voice as per Branded3. So it is not just the mobile devices we are talking about here. By 2020 50% of all searches will be voice searches reports Comscore. The convenience of speech over typing seems to be taking over well and faster than I had imagined. You obviously can search much faster using voice than typing the same. Hence also receive quicker answers for your search.
Imagine carrying shopping bags and searching for a place to eat. You just have to light up your smartphone and speak “Hey Google, which is the best restaurant near me”. If you are using a smartphone having a voice search makes life a tad bit easier. When you do not have to type and instead talk to your device in a natural language you have obviously chosen the more convenient option. Is technology making us lazy? Well, that calls for a whole other debate. So, anyway.
Voice search interface
When we talk about voice search, several interfaces come to existence. Mobile, desktop and AI-powered smart speaker voice search. Andrew Ng made a prediction via Twitter saying:
As speech-recognition accuracy goes from 95% to 99%, we’ll go from barely using it to using all the time! aka ~ Andrew Ng
So, Andrew Ng believes that the accuracy of recognition is what will take voice search into the mainstream. 95% word recognition is actually the same threshold of accuracy as human speech (Google officially reached this threshold last year, to great excitement). Thus Andrew Ng is holding machines to a higher standard than humans. The assumption is reasonable since we tend to address new technology and machine interfaces with a higher degree of uncertainty, giving it little margin for error.
For voice search to become the eventual mainstream search I think we need to go much beyond just precise vocal recognition.
User experience of voice search
….The internet is filled with such funny failure examples.
Using natural language search is more conversational. We tend to have the same expectation that we have when conversing with human beings. In a conversation, we expect machines as well to respond to us in a human manner. Effortless, intuitive two-way conversation. When Alexa or Siri or Google or Cortana or Bixby responds to you with a “Sorry I do not understand your question”, you often tend to be disappointed.
Often, understanding you still is not enough to produce the right response; many voice commands depend on specific phrasing to activate. Thus you can get wrong responses if you don’t know what exactly to speak to obtain the desired results.
Now, while typing you can have typos and always correct it immediately. Or sometimes you can also have Google correcting them for you. But in case of voice search, if you mispronounce a certain word in a sentence, there is no going back. Wait for the device to come back with irrelevant responses until you can correct your question and ask again.
Thus when we say voice search is faster than typing, well is it really yet?
However, such technological impediments as well know will only improve with time. Hence to say none of them is a deal-breaker.
The keyword factor
Bing search engineers have found that voice search keywords are significantly longer when compared to text-based search. For example, let’s say when you want to read a book review, you might type “book review Kite Runner”. But when you end up talking to a voice interface you will say “Hey Alexa, what are the reviews of the book Kite Runner?” Thus as we notice here the queries are completely different. The question now; is voice search now still capable of navigating the web to find the absolute right answer for us?
Google is slowly changing from a search engine to an “answer engine”. In fact, thanks to SERP features like Knowledge Graph and Featured Snippets, the number of organic clicks have dropped by 37%. Why? You don’t need to visit a site to get your answer. It’s right there in the search results.
Google is using this same answer-focused technology to provide voice search results. For example, let’s say you wanted to know how many calories are in a banana. Earlier you would enter a query say ” banana calories” and Google will list you 10 blue links to choose from. But now a featured snippet in the very beginning will give you your answer right away. Sparing you the trouble of browsing through multiple links. You are still provided with the blue links in case you want further information.
The voice search in Google does nothing but reads out the featured snippet to you as well. Infact Google in order to respond to your searches finds a keyword anywhere in a piece of content. If you want to make sure your website is ranked higher, make sure to use natural languages as keywords. Also since it is conversational language do not avoid long tail keywords.
Do not interrelate voice search with AI digital assistance
When I talk about search I am talking about a voice interface having the ability to respond to my query or question. When I ask Alexa to switch on the lights or ask Google to wake me up at 4 am, it does not qualify the criteria of a “voice search”. A query can be asked and answered using any voice interface or call it a machine capable of voice input and output. But commanding a device to do a certain action does not categorise as a voice search entity.
Voice search marketing pathway
Basing on the fact that 20% of searches in the Google app are now by voice, you can assume that voice search is on the rise in the marketing world. Your mobile footprint is crucial to keep your business competitive. No matter how well-designed it may be, your website will not get you the results you want if it suffers from slow load times; website speed is a more important SEO aspect in voice search. Even a two-second difference in load time can increase user abandonment by 32%, resulting in a steady fall of conversions per each second delay.
Another major component of voice search is its ability to “learn” your habits—the way you speak and what search practices. This isn’t just about your accent or vocal characteristics; digital assistant devices also track your search habits. Owing to personalisation you must be aware of how you are bombarded with ads of a product you thought of buying and searched for it on the internet but did not end up buying eventually. This form of retargeting is making its way into voice search as well.
Google stated that their goal is to make their technology omnipresent and seamless, meaning all of our devices—phones, watches, digital assistants, etc.—will have a unified software solution. While this is may be a concerning prospect to some, it is surely becoming a truth in the digital marketing landscape and beyond. To be able to keep up with such a competitive market, your business needs to be on top of the latest technological advances like search engine algorithm updates. Continue to accommodate your SEO strategy according to these developments.
Future…
The voice search if it remains a single result tool should be able to give you the most accurate answer. Or voice search develops into a reliable tool for searching the web. This might involve a user being initially read the top result for their search, and then being presented with the option to hear more search results – perhaps three or four.
If they then want to hear content from one of the results, they can direct the voice assistant to navigate to that webpage, and then proceed to listen to an audio version of the news article, blog post, Wikipedia page, or any other website that they have preferred.