Watch any old sci-fi film or TV show, and a few well-worn tropes will appear. The characters often have flying cars, humans invariably colonise other planets, and technology can always be controlled by your voice. While flying cars and far-flung space colonies are yet to emerge, voice-controlled tech has very much become a reality.
Smart speakers and home hubs like Amazon’s Echo are now commonplace. Almost every smartphone now has an embedded virtual assistant like Siri or Alexa, waiting to answer questions in reassuring (usually feminine) tones. We really can speak to our devices to make them work for us, and one of the most developed features in this regard is voice search. It’s becoming popular with web users, as well as transforming search results for everyone else…
Speaking with one voice
Voice searches involve speaking to a device rather than typing a search term or keyword into it. Speech recognition technology attempts to understand what you say, and then delivers results based on that interpretation. This may involve a search results page, or even answering your query in an electronic voice.
Voice recognition technology isn’t particularly new – there have been speech-to-text programs and dictation software for many years. The explosion in voice search stems from the improved accuracy of speech recognition. Over the last couple of years alone, Google has made great strides in this regard, currently boasting a word recognition accuracy rate of well over 90 per cent. And increased accuracy makes speech recognition tech increasingly useful.
Features become more popular and firms work harder to improve their offerings – a self-fulfilling cycle which suggests searching by voice is here to stay. But what does all this mean in terms of search engines and the results they show us?
Search and ye shall find
There are a number of reasons why the popularity of searching by voice is revolutionising search engine results:
1. The language of voice search is different.
When you perform a traditional online search, you often type a particular phrase into a search engine, such as ‘best cat food’ or ‘top London restaurants’. The language of voice searches is demonstrably different. People asking a question orally will do exactly that. They might say ‘which is the best brand of cat food?’ or ‘where’s the best restaurant in London?’ The results displayed reflect how the search is phrased. They’re dominated by pages that are optimised around more conversational phrases, as opposed to pages designed to rank for traditional, typed keywords.
2. Succinct summaries are king.
The average voice search result is around 30 words long. That’s the kind of result a smart speaker or virtual assistant might read out to you, as an answer to a question. If a page is to perform well as a voice search result, it needs to provide succinct answers to key questions. Longer content tends to rank well in traditional search results, but it isn’t always great at offering abbreviated one-sentence answers.
3. Local mobile-friendly content is more compatible.
Most voice searches are performed on smartphones, which means sites that are mobile-friendly are more likely to rank well than those that aren’t. The searches are also more likely to be navigation or location-based. For instance, people might ask ‘what restaurants are near me?’. AI virtual assistants perform these kinds of searches with a user’s location in mind, meaning sites which integrate with Google Maps tend to return as high ranking results, as do sites whose content has been optimised with local references.