While it still seems like a way off, a new era of voice and image-enabled search is getting closer, with more people turning to voice search for everyday queries, and more people using tools, like Pinterest's Lens, to identify items based on a photo.
That, as you can see in the lower notes, was three years ago, and the increased adoption of smart home devices and related tools has no doubt seen voice search usage increase since then.
But still, voice search remains hit and miss. Sure, it might be pretty good when listening to a single voice in an isolated space - like in your home or in your car - but with added background noise, other people speaking, etc., that accuracy drops significantly.
So will voice really become a significant SEO consideration? And how can you optimize for it to ensure you still get your content in front of the maximum amount of interested consumers?
Definitely, Google's tools, in particular, are improving. The company recently outlined a new speech recognition project which has reduced error rates by over 20%, while Google's new on-device speech recognizer within Gboard recognizes speech faster, presenting output words character-by-character, "just as if someone was typing out what you say in real-time".
So Google's speech recognition tools are getting faster, and more accurate - and this week, Android Police has also reported that Google Podcasts is now automatically generating transcripts of episodes, and using that text output as metadata to help listeners search for shows.
Immediately, you can start to see the broadening scope of speech to text potential - now, you can search through podcasts to find relevant episodes, even without full context of the name of the show or guest.
The implications of these improvements, as noted, are significant, and could lead to a whole new consideration for SEO. Optimizing for voice search is still not a defined practice, but measures like improving site load times and creating a voice-aligned FAQ resource can help in maximizing your potential to be found by those who go searching for related terms via voice tools.
It's interesting to consider the implications here. It doesn't feel like we're close to that '50% of all searches' mark just yet, but it's clear that voice and image search tools are evolving, they are developing to the point that they have practical, immediate value for consumers, and as such, will become a larger business consideration.
If you've not considered the potential of voice queries, the time is coming where it will be a more relevant tool. It's worth looking at how people are using voice, and how your website relates, as the technology advances.