Voice Search Queries: Testing and Refining (Fourth in a Series)
December 26, 2017 by Rob Andrews
Previously, in our series of posts discussing voice search and voice search queries, we have thought about the full scope and impact of voice search, carefully considering the overall characteristics of voice search and how it differs from traditional internet searching. In our second article, we considered how to fully leverage the power of your website, and preparing it for a voice search world. In our last post, we presented some strategies for brainstorming possible types of voice search queries that could be applicable to your niche.
In today’s installment, we are going to get into some of the nuts-and-bolts of how you can test the possible voice search queries that you may have brainstormed.
Voice Search Query Testing: Multi-Platform and Multi-Search Engines
To begin with, one basic principle: Full testing of voice search queries will require you to run different queries on a variety of devices (laptops – desktops – smartphones (both iOS and Android-based) – tablet devices – stand-alone devices) as well testing on different search engines and assistants (Google, Siri, Cortana, Bing). You may not have direct access to all of those devices and platforms, so it’s OK to reach out to staff or even family and friends who might be able to assist you with this. The more variation you can bring into this, the more comprehensive your testing will be, overall.
Documentation With Voice Search Query Testing: What’s Needed
As you “experiment” with different devices and platforms, you need to document the following types of information:
Results Generated: When you run a voice search query, note carefully what results you are getting, including specific site information being used as the source for any proposed “answers.”
Maps: Do the results that you are seeing include any map listings of actual locations? What is being pulled up and what site/business correlates with this information? Could you rank for that instead?
Data Sources: Where, overall, does the data currently seem to be pulled from? Public sites (like Wikipedia) or private sites? Forums or industry portals? News sites? Question/answer sites? Competitor sites?
Paid: When you run a voice search query, are paid ads showing up near the displayed results? If so, what type? How many? For whom?
Single/Multi Listings: When the voice search query is run, is a single result displayed (or spoken) or are you given a number of options that you can choose from?
What To Do If Your Competition Shows Up In Your Targeted Voice Search Queries
Be prepared for it – your competitor(s) may currently be showing up for various voice search queries, when you are not. But this is not the disaster that you might think it is – in fact, it is very helpful!
What you need to be asking yourself (in a situation like this) is: What is my competitor doing that I am not doing, that caused their site to appear in the returned query? In most cases, you will most likely need to modify your existing content, create new content, or change your site structure, coding and/or ranking. A careful analysis of their site is what needs to be done here. If they can get there, in most cases you can as well.
Brand Searching In Voice Search
Related to this is another phenomenon: When you run test queries for voice search, having to do with your business brand/product/service, the returned result(s) may not show your brand but your competitors. This is a strong indication that your competition is out-optimizing you for your own brand/product/service – currently, they are piggy-backing your search traffic for their own benefit!
Obviously, that’s a situation that you should not settle for, so again, you will need to carefully consider, through analysis, what it is that your competitor is doing that you are not, and set out to correct that.
Does Your Site Currently Have Content That Can Rank For A Question?
As we have mentioned in our (link) previous article, the fundamental basis of most voice searching is questions – searches that are phrased in the form of a question, and so your site needs to address those specific questions. But perhaps it already has content that could rank for a question.
A very crude estimate (and I’ll explain why it’s just a rough estimate in a moment) of your chances of ranking your existing content for a question can be determined by doing two quick searches:
site:domain.com “search query” (with quotes – exact phrase match)
site:domain.com search query (without quotes – partial phrase match)
These site domain searches against a question will return results where your page content, across your entire domain, has text matching – either partially or completely – a question. This is only a rough estimation as the search engines today are more sophisticated and may actually rank pages in your site for a question that is not explicitly phrased as such, but it’s a handy starting point. Of course, if you identify any such pages, that’s a key page for you to try to get ranked for that question (and perhaps others similar to it).
Keep in mind: pages that you have that could rank for a question may not be the page that you are actually trying to rank. So, in that case, you need to ask which page has the content that best answers the question? Which page would you like to rank for that particular question? Then adjust your overall strategy going forward to target the page that will benefit you the most.
A Couple of Sources of Good Information
You can actually use the AHREFS tool to see if you are currently ranking for Featured Snippets — and here is a great study (by AHREFS) on 2 million Featured Snippets. These two articles can help you refine your voice search testing. And just a side note: at present, Google will not be adding analytics relating to Featured Snippets into Search Console.
In our fifth and final installment in this series, we’ll consider some additional killer ideas to help propel you to the top of the (voice search) heap, so stay tuned!