- About us
- The team
- Contact Us
Our third instalment in the ‘Writing for Voice Search’ series dives into how voice search works. If you’ve missed our previous two posts, catch up here. Today, I’m going to take you on a quick trip through the history of voice search and how the technology actually works. Technophobes, don’t be afraid – we’ll keep it light.
While smartphones and smart speakers might look futuristic and super-advanced, the hardware for voice search has been around for decades. At its most basic, all voice search requires is a microphone, a computer processing unit, access to the internet and a speaker.
Smart speakers and other voice search enabled devices use their microphones to pick up words. Using complex natural language processing systems (the new bit of the technology), they translate what you’re saying into a command. This could be to use a search engine to find an answer, to scan your music library for a certain song or to use an app to find information about the day’s meetings, weather, celebrity gossip … the list goes on.
Apple with Siri, Google with their Google Assistant, Microsoft with Cortana and Amazon with Alexa have been the technology companies flying first into bringing a consumer-facing voice-interface. They want to continue to be the gateway to information – the point where we go to discover knowledge, music, entertainment, events… anything and everything.
By being the gateway, you can control the flow and monetise access. And the use cases for voice search are many, and incredibly lucrative.
Voice Search began in 2010, when Google introduced a revolutionary (albeit clunky) telephone service for searching the internet. All users needed to do was call a number, where they’d be prompted by a recording to ‘say your search keywords’. This would then open a webpage or send a link with answers to the device, which the curious user could then browse.
This was impractical and a far cry from today’s strangely beguiling virtual assistants. However, it did pave the way for research and development in the world of voice search. A few years ago, Google’s ‘Hummingbird’ update changed the game by introducing semantics into the equation. So called ‘semantic searching’ now sought to understand the intention behind each search in a more human way, bringing results that satiated the user’s curiosity rather than just finding select keywords in unrelated articles.
Initially, voice search technology faced teething problems, including issues recognising speech of users with strong regional accents. There has even been debate around whether voice search is contributing to the decline of regional accents, as more than 79% of users admit to softening or dropping their accents altogether when using voice search.
Research by leading pioneers including Apple and Google has improved the technology since it was first introduced, though. For example, Google’s new Gboard, featured on its Pixel phones, promises to be one of the most advanced in the world. It picks up on things like intonation to correctly punctuate sentences and uses context to overcomes grammatical queries (like the there, their, they’re conundrum, which many humans are still yet to master).
However, there is still a way for voice search technology to go, particularly when it’s being used in noisy environments or by those with strong accents or speech impediments. Only once these hurdles are overcome will the technology be truly accessible for those who need it most, like disabled users. There’s also a long way to go for those in the world of communications too, as we re-evaluate SEO best practice in light of the ever evolving way those we’re trying to reach are searching.
Next, we’ll dive into how you should approach writing for voice search in more detail. Because, unsurprisingly, we don’t speak to our phones in the same way as we type with them. Keep your eyes peeled for it over the coming weeks …
This is the third article in our series on voice search. The full list of posts are below: