The rise of voice search is no secret, and many companies are still wondering how to address it. When it comes to search data, how can we monitor which queries are from voice? In this article, Jason Tabeling shows how he finds insights into voice search from his own company’s data.

Alpine.AI estimates that there will be more than 1 billion voice searches completed in 2018[1].

At this point I’m sure that everyone has seen or done a voice search, even if you just saw an example in a Google Home or Amazon Alexa ad. The power of what can be completed with a voice command is growing by the day. This trend is already having a massive impact on consumer behavior and therefore needs to be a consideration when monitoring or optimizing our search accounts.  

Right now Google doesn’t provide specific information on how a search was started. For example, was it a Google home search, from the app, done via typing or voice.

However, I wanted to take a recent dive into our own search data from our company, BrandMuscle. I used Google Ads client data across all verticals, comparing the first nine months of last year with this year, dates 1/1/17-9/30/17 versus 1/1/18 – 9/30/18.

The data revealed some interesting trends that give insight into how voice searches might be showing up in the data we currently have available.

Looking at the length and type of queries

To start, I looked at the length of search queries. The thought is with voice search consumers are using a more natural language. So instead of searching for “car insurance” consumers might search for “cheapest car insurance for Toyota Camry.” Our data shows the average number of

Read more from our friends at Search Engine Watch