The Oban Blog



A fast-changing world: The inexorable rise of voice search

September 21, 2022 Digital channels

When you look around, do you see people talking to their phones and voice assistants?

How you answer will depend on where you are. Not just your immediate environment – for example, at home, in the office, walking down the street or in your car – but where in the world you are. For example, Google recently stated during their Google I/O ’22 keynote that nearly 30% of Hindi queries in India are spoken – but this proportion varies by country.

Even five years ago on a trip to China, I was surprised to see so many people speaking commands into their phones as they walked down the street. Personally, I feel self-conscious doing this – possibly because I’ve been using search engines for about 25 years, and voice search has only really gained traction in more recent years. Before voice, I was so used to typing search queries and while search engines have evolved significantly in the last two decades, much of my searching is still based on the ingrained assumption that search engines need the right phrasing to generate the right result.

Of course, as an SEO, I know the reality is that at least Google and Baidu are becoming increasingly sophisticated in understanding conversational search. MUM for Google and Ernie 3.0 for Baidu mean fully fledged semantic search based on machine learning that applies to understanding the potential results as well as understanding the search query semantically (which was the limitation of previous updates such as BERT).

For those internet users who have only started using search recently, naturally their search habits will be different. Younger people will search differently although they will still be influenced by the culture they grow up in and possibly their elders.

In countries where searching the internet is more recent for everyone and voice and semantic search has been there from the start, voice search is a more natural, intuitive way to search – allowing you to search conversationally and receive the results you are after.

 

How search evolved to become more conversational

Despite suggesting that some of us who have used search engines since the beginning are sometimes stuck in our ways, our collective search behaviour has evolved as search engines have improved.

When online search was relatively new, anything other than a broad topic was unlikely to return a relevant result. We would search the likes of Alta Vista or Lycos often for a type of site and then find what we wanted using that site’s navigation. For example, back in 1998, I might have searched for ‘video stores’ on Alta Vista and found a store where I could then search for the VHS I wanted to buy.

Ask Jeeves claimed in their ads at the time that you could ask a question and Ask Jeeves would find you a clear answer, but they never seemed able to deliver on this promise.

In the arms race to increase market share, search did slowly improve but when Google came along it took a giant leap forward, although early Google would seem painful to use by comparison to today.

By 2003, I could search for ‘Finding Nemo VHS’  and many of the results would be relevant to me wanting to buy. However, if we wanted to research and find answers to specific questions, we would still struggle. It wasn’t impossible but it would mean using just the right words, perhaps trying a dozen combinations to find the one that yielded the result we were looking for.

There was clearly demand to make search easier and more relevant. By the late 00s, Google dominated the market and was refining the context of search queries. Progress was incremental so that as users, we often didn’t notice the evolution in search result quality or the impact on how we searched.

Users definitely noticed the major changes that took place in the mid to late 2010s, with Google finally being able to deliver on what Ask Jeeves promised by providing a featured snippet that directly answered, most of the time, specific questions, or research intent terms. This functionality soon became used for voice results too.

Soon after, Google introduced BERT to improve understanding of the semantic meaning of a search term, improving the quality of results almost immediately. Now, searching takes much less time, is more intuitive and more accurate, with a bigger pool of results that Google will realise are relevant even where the exact keywords aren’t found.

 

Barriers to voice search

According to Google, about 27% of users around the world use voice search on their mobile. A 2020 study found 73% of respondents blamed accuracy for the reason they didn’t use voice, while 66% mentioned accent or dialect related issues and 38% cited language coverage (or lack thereof).

The quality of search results varies by language, because of the number of available results and the fact that machine learning relies on multiple sources. Often this means users in certain languages still search more broadly to find the results they need. With Google now translating results in real time for a number of languages such as Indonesian and Hindi, search quality could see a jump, as could conversational voice search usage.

Improvements will keep coming, which means people will increasingly talk to search engines more conversationally. And in fact, data from 2019 showed that Google Assistant understood 100% of questions and answered 93% correctly of the time.

 

Oban can help you find your voice

Sceptics may think that talking to a device in a conversational way is something that will only appeal to a tech-savvy minority. But when we look at how far we’ve come, it seems likely that we will continue to evolve our searching habits, although perhaps some markets and audience groups faster than others. Voice search is here to stay, which is why search marketers are focusing on how to use it both now and in the future.

Optimising your content strategy for voice search doesn’t need to be a separate and standalone activity. Most practices which help voice search optimisation pay SEO and UX dividends, and – as they involve deepening your audience understanding – are absolutely worth doing. To find out how Oban can help you find your voice across markets, please get in touch.


Oban International is the digital marketing agency specialising in international expansion. Our LIME (Local In-Market Expert) Network provides up to date cultural input and insights from over 80 markets around the world, helping clients realise the best marketing opportunities and avoid the costliest mistakes.