The marketing industry is all a buzz about the potential of voice search.
Estimates are that the number of Voice Assistants in use are predicted to jump threefold to 8 billion by 2023 from 2.5 billion. And those voice assistants are being used to search for answers/information via voice (vs executing a voice action).
Bing surveyed voice users and found that 72% of them have used voice search to look for information through a personal digital assistant (Siri, Alexa, Google Assistant, Cortana). Bright Local’s 2018 study research saw that 58% of consumers have used local search to find local business information in the last year.
While those numbers are compelling, how do you actually make a case to your boss or client that YOUR audience is using voice search and therefore you need to create a voice search optimization plan?
And once you get approval to launch a voice SEO effort, is it possible to measure the ROI of that program?
You’re going to find that measuring voice activity is challenging. This post will examine what data you can use to make the case for a voice SEO program and outline the various challenges that exist currently to measure the ROI from that program.
A few tips for making the case for investing in Voice SEO
There are five major data sets you can look at to make the case. I will walk you through how to use research data from your target audience/demographic, data you own (Google Search Console, Bing Webmaster Tools, and Google Analytics) and data from an SEO keyword tool to make your case and develop your plan.
Let’s get to it.
1. Research data related to your target audience’s use of voice.
Does your business target one of the demographics that are statistically more likely to use voice search more frequently?
It turns out some demographics are adopting voice faster than others (shocking I know). Here’s the most recent survey data related to voice and age demographics from Path Interactive:
Key takeaway: 50% of users ages 65-plus claim to use voice search 1-3 times per day!
Or, voice search optimization should be a higher priority for you if you have a business where people are more inclined to search via voice, like local businesses, or e-commerce – for instance, 15% of voice speaker users are engaged in purchasing via voice:
Demographics and survey data related to consumer behaviors alone might be enough to make your case. But what if your business is not a local business and does not sell products online?
Let’s try the next data source.
2. Web analytics data.
While that seems like it should be a no-brainer, you, unfortunately, can’t measure how much voice search traffic is coming in via Bing or Google via web analytics.
Clicks from most mobile search widgets and apps (which include those with voice search) are recorded in Google Analytics as visits without any attribution are being dumped into the “direct” channel (vs. “organic). Visits from Apple iOS Safari (after version 6) also show up as direct vs organic. I wrote about this web analytics issue earlier this year.
All visits from Google voice search sources (Assistant, Home, and Google Siri) and voice searches from Bing-powered devices like Alexa and Cortana are also recorded as “direct”.
So while this is where you would be tempted to check, it won’t help you.
Keep the above in mind if you were wondering why your direct traffic was increasing or conversely if you feel like your organic traffic isn’t increasing as much as you hoped.
You *can* get a sense for how much of this “dark traffic” is actually organic by going into Google Analytics, selecting the “direct traffic” segment and looking at the all pages report. Most likely any page in that report (outside of your homepage and other pages users might type in directly or bookmark) is either a page with lost GA tracking or traffic from one of the above sources.
3. Voice keyword volume in Bing or Google Webmaster Tools.
What you can do is guess based on the queries you are targeting which look like voice search queries – more natural language type queries in Bing Webmaster Tools and Google Search Console.
Side note: Mentally I have issues with this approach as these queries could also be autocompleted by the search engine when someone is typing.
In Bing, look at the Search Keywords report. Look for queries that could be voice queries (i.e. they are more natural language queries), like this:
In Google Search Console, you’ll find the information under “Search Results”.
If you are using Google Search Console data to make the case to your clients/boss, be aware of how Google Search Console tracks rankings and what data is not being tracked.
4. Keywords targets that triggering Featured Snippets.
Determining how much voice traffic your site is getting is undoubtedly a challenge. In many cases, the ranking Featured Snippet is used as the voice answer. So at a bare minimum, I recommend to all my clients you start tracking your Featured Snippet and PAA rankings (in Bing and Google) with your SEO tool of choice. (GetStat, SEMRush, and Ahrefs all provided this tracking).
Also, if your site is not currently ranking for terms that you think are voice queries and you rank for that term in the top 3-5, you can work on modifying your copy to rank for that top slot. I have a Featured Snippet guide here. I would also recommend reading Google’s guidelines on what content they think is high quality to be read to Google Assistant users.
But – sometimes the Featured Snippet is not the voice answer.
I stumbled onto great research where the researchers used GetStat plus some manual research to see whether the result presented by Google Assistant for a voice query was the same as the top Featured Snippet, or something different.
For some queries, Google is partially relying on Knowledge Graph information for the answers, and in some cases will show a well-optimized Google Action.
So you need to do some work to ensure that your brand and products are part of both search engine’s Knowledge Graphs.
Ensuring that your information is a part of the Knowledge Graph involves ensuring that your brand and products are appropriately mentioned in databases that Google/Bing uses for the Knowledge Graph (Wikipedia, Wikidata) as well as marking your site up with schema.org markup. I’ve covered more in-depth information about the Knowledge graph in this post.
5. Data from testing to what is the voice answer for your target topics.
It makes the most sense to just test to see what the voice answer is on the various devices for the topics for which you want to appear.
But can you scale this activity?
There is a Voice Readiness Tool provided by Synup which gives you an approximation from Google Assistant and Alexa, but it only works for local businesses, not a product or general information queries.
Here’s an example from when I searched for a local yoga studio:
For non-local queries, I think you would just have to test manually. Let me know in the comments if you know of a tool I’ve missed that would speed up this process.
Measuring your voice SEO program.
So let’s say you convinced your boss or client to start a voice SEO program. First, congrats. But you know that you’ll need to measure your return on investment (ROI) to justify your spending time on the effort.
Proving voice SEO ROI is challenging as the SEO tools on the market don’t (yet) provide this feature. All of the major SEO tools measure Featured Snippets but don’t provide voice ranking.
Also, most SEO tools are desktop-oriented. Their mobile SERP tracking doesn’t capture all of the features in mobile SERPs, which highly skews what a #1 ranking might actually look like. None of them measure or report on the additional features that could be displayed in a Featured Snippet that might impact click-through rate.
Here’s an example. Here’s what the mobile SERP looks like:
And here’s how Ahrefs displays the mobile SERP:
Did you notice that the Featured Snippet had tabs for “Overview”, “Videos” and “People Also Search for”?
Those are features that are NOT mentioned in the Ahref ranking report. And look at all of those dropdowns! These mobile SERP features also change fast. While the screenshot above is only from a few weeks ago, now the Featured Snippet has a “top stories” and a video carousel in it – as a part of that Featured Snippet block.
And if you flip tools, it doesn’t get much better. While SEMRush will give you an actual mobile SERP snapshot, often it’s months out of date, and so it’s not really actionable.
And none of the SEO software tools report on actual audio rankings – what was read out at for the query to the user. To measure that, I would assume you’d have to test yourself on the various devices, though they would over time become personalized to you and your history, which would skew your results.
I am a proponent of making sure that my clients are future thinking and are actively adjusting their online marketing plans for new user behavior — and I AM working on getting them to plan and optimize for voice queries. But you can see why I think getting clients excited about voice SEO (if they are outside of the e-commerce/local search verticals) is tough.
How did you make the case for voice SEO for your voice or clients? And how are you measuring the ROI?
Let me know in the comments below.