Voice search is the future. We all know it. We’d all much rather have a conversation with an intelligent device than initiate a lifeless keyword query.
What’s not so clear is how marketers are going to deal with voice search. A long time ago, when it came to search engine optimization, keywords were king. Then some years into its evolution, content became king, then it was incoming links, and so on. What is it going to be now that voice search is here?
Why voice search is taking over
Voice-assisted search came onto the mass market in 2011 with the initial release of Siri. ComScore predicts that 50 percent of searches will be done through voice by 2020. Additionally, Gardner predicts that 30 percent of web browsing will be done without a screen.
A survey done by MindMeld found that more than half of respondents had started using voice search in the past 12 months. Today, 20% of queries on Google’s mobile app and on Android devices are voice searches.
With voice recognition accuracy rising at an exponential rate, it’s only a matter of a few years before search engines are mostly replaced by so-called “smart speaker” devices such as Google Home and Amazon Echo which are completely voice controlled and don't include any other kind of interface. (Amazon has sold about 10 million Echo smart speakers since its 2014 introduction.)
Fifteen years ago, the error rate of voice recognition was 20-30%. Today, the average speech recognition word error rate is a drastically lower 8%. Artificial intelligence is now learning how to improve itself autonomously. Microsoft’s latest system, which has six neural networks running in parallel, has reached an impressive 5.9% error rate (which is about the same as a human transcriber). Soon, voice recognition will likely be better than other humans and we will be able to hold complex conversations with our own devices.
Today, consumers can ask their virtual assistants such as Siri, Alexa, and Cortana to check the weather, sports scores, flight status, look up factoids, do math, call an Uber, order a pizza, etc. That list is going to expand to do just about anything you can imagine a consumer would need.
Shopping is also a major part of voice search as well. In the last 12 months, 19% of consumers have made a purchase using a voice-controlled device. Amazon Alexa, for example, allows users to order a pizza from Domino’s, or hail an Uber driver. Google also announced that it has already integrated a range of third-party transactional services into Google Home, including Uber, Pandora, OpenTable, Spotify, WhatsApp and Ticketmaster.
Voice search statistics and predictions
The rate at which voice search is expanding is exponential. It’s increasing faster and faster every year. It’s only a matter of a few years before most consumer queries are done with natural language. Here are some stats to back that up.
- “50% of all searches will be voice searches by 2020.” — Comscore
- “About 30% of searches will be done without a screen by 2020.” — Mediapos
- “40% of adults now use voice search once per day.” — Location World
- “Cortana now has 133 million monthly users.” — Tech Radar
- “25% of 16-24 year-olds use voice search on mobile.” — Global Web Index
- “11% of people using voice search started more than 3 years ago. 41% of people using voice search have only started in the last 6 months. 60% of people using voice search have started in the last year.” — MindMeld
- “325.8 million people used voice control in the past month.” — Global Web Index
- “42 percent say voice-activated devices have quickly become “essential” to their lives.” — Geomarketing.com
- “Nearly 50% of people are now using voice search when researching products.” — Social Media Today
What does the emergence of voice search mean to marketers?
With AI and voice recognition, we are nearing the end of the road when it comes to the evolution of search engine optimization (SEO) best practices. In the not-too-distant future, artificial intelligence algorithms such as deep learning will make it easy for search engines to provide better results than any human being regardless of keywords, content, incoming links, or any other trick you might want to pull out of your hat.
AI bots will develop strategies based on big data and deep learning that human minds may not even be able to comprehend. Such is the case with Deep Mind’s AlphaGo computer which, after only a few days of playing the game against itself, developed strategies that no human is familiar with. AlphaGo is now the world’s best Go player, mastering a game that relies as much on intuation as it does sheer strategy. The same will happen with search results.
The end goal of the evolution of search technology is not to find better search results, it’s to find the best answer—not an answer that will please everyone, but an answer custom-tailored for the person making the query. AI advances in both voice recognition and search results will produce the perfect personal assistant not too far in the future.
Optimizing for voice search
Already, 60 percent of searches currently take place on mobile devices. One need only look at Bing’s market share to see the power of mobile search. As a result of all the Android and Google-powered voice searches that utilize Bing’s data, their market share has rebounded substantially, especially in the past couple years.
Unlike desktop Googling, voice searches don’t deliver a long list of possible search results to scroll through. They generally provide the top one or two potential answers to a query. This makes the top one or two places in search results worth their virtual weight in gold.
Voice searches take a more conversational approach than keyword searches and are much more specific. Users typically ask a long question to a device where they might type a short string of keywords in a screen based search. For example, in the old days, a typed query might be, “China vs. India tea prices.” With voice search, this becomes “What’s the difference in the prices of tea in China and India?”
We’ll leave you with a few solid tips for improving your chances of being included in the results of a voice search query.
- Don’t just optimize for keywords, optimize for questions your customers are asking. Make one page for each question which could serve as the top search results. When you ask Siri a question and she says, “I found this on the web.” You want that to be your page. Name the page with the question.
- Include search terms like “best” or “top” in your questions (and page title) which were previously undervalued by search engines, as these tend to be used a lot in voice searches. For example, “Where are the best planetariums in the US,” or “What are the top five most popular dog breeds?”
- If you offer something that can be purchased or scheduled right then and there, include a timely call to action with an immediate click action such as, “Book a museum tour now,” or “Call now to make an appointment,” to take advantage of the mobile mindset and the perceived immediacy of the need.
Someday we’ll all be carrying on conversations with our computers. And we’ll all have a direct line to some cloud technology along the lines of IBM’s Watson which can answer questions in great detail and even take care of all our planning, reservations, purchases, etc.
We’re still not quite there yet, but brands must begin to adopt better strategies than jamming in keywords or building tons of incoming links because most of these strategies will be useless in only a few years time. Brands need to focus more and more on simply being great at what they do or offer, providing consumers with answers to actual questions, creating quality content, and making it easy for smart speakers and virtual assistants to interface with your web properties.
For more great digital marketing tips, check out our article, “2017 Digital Marketing Checklist.”