Some verticals are more suited to voice search than others — but which ones? John Campbell of ROAST has the answer, and the data, for you.
Watch John dig into the nuances between verticals with a bite-sized version of his talk from Manchester City Crawl and learn how best to optimize for voice search and stay competitive in your space.
Video transcription
Hi, my name’s John Campbell. I’m the head of SEO at ROAST.
So, today I’m speaking about voice search. We’ve done a study into 22 different verticals and we’ve looked into how Google answers questions in those 22 different verticals.
The ROAST voice search report
The ranking report that we conducted looked at 10,000 different key phrases and we ran it in April, all from our office in London.
The key phrases were a split of generic key phrases and also questions that people might have in their specific industry.
We pulled that information back, categorized it, and at the same time we pulled back STAT data with featured snippets on the same days, and merged those two pieces of information together. For each vertical, we’ve written a whitepaper, which goes into a little bit more in-depth into the findings.
“We found that voice search does vary according to industry and that it is very much dependent on the key phrases we’re using.”
For a certain industry, such as fashion and transport, we’ve seen the assistant’s not very good at answering questions because the queries are quite complicated.
Then other industries such as insurance, there’s really quite a high level of competition there already, and the comparison websites from the direct insurers, there’s already kind of a battle going on in those search industries.
From the initial report, it was really the kind of insight that we found that Google was using lots of different datasets to answer questions. Originally, we just thought it was just all about feature snippets, but then we soon realized that Google uses its knowledge graph for information.
It could also use information from actions and it could also maybe have pre-set routines that will go through in conversations for things like flight search.
Those are the kind of interesting things we found out and it’s stuff that we never knew before without running the tool.
How to optimize for voice search
Some of the notable things that you should think about when it comes to voice search is that you can gain results obviously by optimizing your website for featured snippets. Also, you might want to think about your Google My Business Listings, and find out if you feature on other people’s featured snippets.
If a magazine has talked about your product, how can you get your product listed into those featured snippets and then get your answers read out. The other one is that Google uses the knowledge graph as well.
If you were the owner of a landmark such as this, of an art gallery, you want to promote it and you’re searching for something like what to do in Manchester, you know, that’s an option for you to promote your knowledge graph result to get into those results when Google reads them out.
Using STAT
In STAT, one of the things we use quite a lot is the “people also ask”, it’s the ability to extract that information from the search results. So, one of the things we do that for our clients is we extract that information and then we create graphs (which is matching up subsequent questions that people have), grouping those questions, finding out where they have results, where they don’t have results and that visualization helps our users see, you know, certain pain-points that customers might have and where they can maybe create content to answer those questions.
Liked John’s talk on voice search?
If it’s voice search you’re strategizing for, then you’ll love reading through their full voice search ranking report, here.
Want to explore the relationship between voice search and featured snippets? We took a deep dive on the influence that voice search has on Google’s favourite SERP feature — and the findings will help you nab more snippets and optimize for voice search at the same time.