This week, I attended a series of conferences on the theme of discoverability, where the presenters tried to explain how to ensure the influence of brands. A short sentence repeated by several speakers challenged me, as I always hear it: “50% of searches on the web will be done via voice search in 2020. “If you work in content marketing or search, you may have already read or heard this sentence at least once. Am I wrong?
Despite my strong desire to launch the debate, I refrained from intervening with the following question: we’re now in 2019, in this room full of web and communication professionals, how many of you do at least one voice search per day? The answer would probably have been well below 50%, and for good reason: the functionalities of voice search are still limited. However, the myth persists. Too many digital and search professionals have been using this data as an argument of urgency and prioritization to clients who eventually drink the same kool-aid.
Please give me a break.
I’m evidently not the first to speak ill of this perception changing statement. A few months ago, while attending the SearchLove Boston Conference, I was pleasantly surprised to hear one of the presenters from Distilled’s shout from the heart: “Please, we are search professionals, let’s stop using this statistic as an industry.” Let’s be smart. A statistic taken out of context is not of much value.
But where does this alarming 50% prediction comes from anyway?
The initial quote was taken from a September 2014 Fast Company interview with Baidu’s former Chief Data Scientist, Andrew Ng:
At the moment, around 10% of Baidu search queries are done by voice, with a much smaller percentage carried out using images. If progress continues at its current rate, however, Ng forecasts that in five years time at least 50% of all searches are going to be either through images or speech.
For the record, Baidu is the most important search engine in China, with an overwhelming market share of 70% in December 2018 (vs. Google’s 2.59%). The statistics were impressive at the time: in two years of functionality life, voice search already represented 10% of the total searches done on the Chinese search engine. But these forecasts need context : they were specific to the Chinese market and Baidu, based on the launch two years earlier of Baidu’s voice assistant, and also including image-based research.
In Google’s world, the first voice search features were released in November 2008, well before Baidu’s. However, it’s only in May 2016 that Google announced that 20% of mobile searches were done via voice. And Google here was specifically talking about mobile, unlike Baidu, which in 2014 was talking about 10% of global voice searches. Why such a difference between Asia and, to generalize, North America and Europe? Consumption habits.
China has experienced incredible growth in recent years, which has changed the way Chinese people consume. Added to this, the cost of smartphones has fallen sharply. These major changes have made the Internet accessible to hundreds of thousands of people who until now could not afford to buy a computer. These new Internet users – or mobinauts – had therefore never learned to use a computer keyboard and thus encouraged the emergence of voice research in China. Andrew Ng mentions in his interview:
The primary catalyst for a step-change in how search works today is the rise of smartphones and tablets, which are taking away more and more market share from traditional PCs. This is particularly evident in countries like Baidu’s birthplace China, where many users are connecting to the Internet for the first time–primarily by way of mobile devices. Of the 632 million Internet users in China as of June this year, 83% accessed the web with a mobile phone, according to figures from China Internet Network Information Center. Most of these users haven’t organically learned how to use text-based search.
So that’s the background story. Besides, after some research, we quickly realize that neither Baidu, nor Google, have recently communicated on the market share of voice search compared to text search. Their priorities are elsewhere, mainly artificial intelligence if you want to know. Speech recognition is still very perfectible (especially in a language other than Chinese or English). This Italian-American grandmother is a living proof of it:
But then, what are the trends in the voice search market?
Better than a few more lines of palaver, I highly recommend reading the articles of Rebecca Sentance who recently gave an excellent overview of the state of voice research and its future.
But before giving some numbers, I would like to address the confusion that seems to prevail among marketers, users and studies between voice assistant (a software agent who can perform tasks or services for an individual, according to Wikipedia) and voice search (allows the user to use a voice command to search the Internet, or a portable device, also according to Wikipedia). The limit is unclear but here are three examples:
“Say Siri, what is the capital of Panama?”
This would be an example of voice search on a search engine (probably Google, since it is the default search engine on iPhone) using a voice assistant.
“Alexa, turn off the kitchen light”
But this, an example of voice command to a virtual assistant, which does not require a web search.
On Google I hit the search bar, click on the microphone (or use Google’s voice assistant) and launch a voice search:
Therefore, voice command and voice search are often found side by side in studies, presentations or pitchs, which is very embarrassing to be able to really know the true percentage of the voice search market share. We all understand that “Alexa / Siri / Google, plays a piece of the Creole Company” has nothing to do with “Alexa / Siri / Google, in which year did the first title of the Creole Company come out? “, one being a command, and the other being a search. Eric Enge, founder of Stone Temple, one of the largest SEO firms in the United States, says it himself:
I also wouldn’t consider using voice to text to message a friend as a voice search. Or setting a timer or alarm on your phone. Yet, personal assistants like Google Assistant, Alexa and Siri do consider these examples of voice usage.
A very interesting study done in 2017 and 2018 (but questionable since it is based only on a sample of 2000 American mobile users) shows several trends:
These trends show that we will be far from 50% of voice searches in 2020, voice commands included. And even if voice search is increasingly used, we are not yet at a generalized use.
The willingness of many SEO experts to aim for the 0 position as the ideal way to respond to voice searches (and that this is what should be done first) is misleading. As a reminder, position 0 is a display format that is added between paid search results (AdWords) and organic search results (usually there are 10), it is also called featured snippet. Here is an example:
Do you ever think that Google’s search results will be limited to a single result (the 0 position) because everyone uses voice search? Me neither. Imagine the SEO war that would result! The 0 position is great for many reasons and it is indeed the one that is read in 80% of cases during a voice search. So what? Is a 200-character response (or a little more in some cases) on subjects as varied as banking, insurance, e-commerce or travel enough to encourage the user to go further in his journey on the same site after a voice search? No. At most, it will generate notoriety. Did you know that a 0 position only generates an average of 8.6% clicks compared to 19.6% for a 1 position (or even 26% when there is no 0 position)?
We are still a long way from the time when we will be shopping for a mortgage or a trip to the South through voice research. While clearly users do not use voice search for these, they are way more likely to use it for very specific questions (calculation, cooking, weather, directions, general knowledge). Optimistically, they could use it to make purchases in the near future, but Alexa’s case shows that almost no one converts via a voice assistant (and those who do will not try again in 90% cases). So, where are we going with the voice search?
Early last year, Eric Enge was conducting an interview with Duane Forrester and Brent Csutoras on how voice terminals will affect the future of search. In spite of the fact that everyone agrees that the voice is more and more present, it’s still very difficult to see where it’s going.
What was the last voice search you did? Do you get angry with your personal voice assistant?