Get ranked.
Grow your business.

Find out what's holding your business back in search?

Voice Search SEO – Structured Data, Entities, & Intent

Voice search, structured data, & artificial intelligence (AI) is the future, and the future is now.

Voice search is growing rapidly, especially among B2B search queries and teens.  But on an aggregate scale, voice search is arguably the most important trend as you look to the future of search and how people interact with things such as mobile devices, cars, appliances… pretty much everything, including each other.

Here’s a graph that shows the growth rate of voice search:

Voice Search SEO

Below are some quick data points to illustrate the rise of voice search:

Stats on voice search, mobile, and AI:

B2B Voice Search

According to BCG, mobile search contributes twice the ROI for B2B transactions compared to searches made from a desktop or laptop.  Here’s a key quote from BCG’s research on B2B mobile search:

Because the effect of mobile marketing ­extends beyond the smartphone to the desktop or laptop, tablet, and offline sales, it’s important that this testing include these other channels. For example, the e-commerce division of one large industrial company was seeking to drive higher marketing ROI and determine the right amount of spending to dedicate to mobile. Using a regression-based approach to ­estimate mobile search’s contribution to overall revenue growth, including conversions that took place in other channels, the company found that ROI from mobile search was twice that from desktop and laptop paid search, after controlling for other factors. Because mobile search was making significant contributions to increasing sales, the company could increase its mobile spending and continue to improve overall ROI.

Along with the fact that 50 percent of all searches will be voice search by 2020, and 50 percent of B2B searches are made via smartphones, it’s clear that if you target the B2B market, it’s imperative that you shift to a voice search SEO strategy for your business.

Entities, Schema, & Structured Data:  Optimizing for Voice Search

In order to optimize for voice search, it’s imperative that you leverage Schema markup (JSON-LD) within your website.  Along with marking up the key pages within your website, it’s equally important that your business data is distributed across the most authoritative sites on the web and that your data is accurate and consistent.

For local SEO voice search, your business data must be consistent and accurate as it relates directly to your Google My Business listing (GMB / Google Maps listing).

For example, if you are implementing a franchise SEO strategy, aligning your business data with your location’s GMB listings, and then distributing that data (otherwise known as citations) is key to getting found in local search results, maps, and apps via voice search.

According to Google, JSON-LD is the recommended format for implementing structured data into your website.

Here’s a simple example of company describing their contact information via JSON-LD / Schema markup:

<script type="application/ld+json">
  "@context": "",
  "@type": "Organization",
  "url": "",
  "name": "Unlimited Ball Bearings Corp.",
  "contactPoint": {
    "@type": "ContactPoint",
    "telephone": "+1-401-555-1212",
    "contactType": "Customer service"

The entities in the example above would be data points such as:

  • Type:  Organization
  • Name:  Unlimited Ball Bearings Corp
  • ContactPoint / Telephone:  +1-401-555-1212
  • ContactType:  Customer Service

There are currently 16 core types of content (entities) that can be marked up with Schema / JSON-LD, according to Google’s structured data guide:

  1. Articles
  2. Books
  3. Courses
  4. Datasets
  5. Events
  6. Fact Check
  7. Job Postings
  8. Local Business
  9. Music
  10. Paywalled content
  11. Podcasts
  12. Products
  13. Recipes
  14. Reviews
  15. TV & Movies
  16. Videos

And there are a few beta features that Google is testing for marking up for search results such as:

Software apps

Apps Schema JSON Voice Search

Top Lists

Top Lists Schema JSON Voice Search

Live Coverage

Live Coverage Schema JSON Voice Search

The reality is, structured data drives intelligent search.

And intelligent search doesn’t imply some kind of super smart user that’s searching in really sophisticated ways.

Smart search is on the UI / AI side of the equation… the technology side.

Voice search is at its best when someone simply speaks into a device and gets the best response or action desired.

Structured data and SEO

When pages include structured data, this means the page is marked up with properly formatted Schema language.  Google prefers the use of JSON-LD as the language for implementing Schema into a page and marking up content.

Without Schema markup, or structured data with the page, Google and other search engines and apps, simply see a page of text.

In this case, search engines have to rely on a much smaller segment of meta data to truly understand what the content is about such as the title and description tags, or image alt tags.

However, this meta data accounts for a very small percentage of the overall content and context of the topics on the page.

Structured data allows the brand to provide a much richer set of data points relative to the topics covered in the content.

Taking structured data one step further, you can also create associations across the pages of your website to provide even more relevance and structure to your pages and content.

For example, let’s say you are a personal injury lawyer in Chicago with 3 partners and 25 associate lawyers.  Here’s a great strategy for implementing structured data:

  • Homepage (organization / Schema type:  Attorney)
  • About Us page
  • Partner attorney profile 1 (partner 1 / Schema type:  person)
  • Partner attorney profile 2 (partner 2 / Schema type:  person)
  • Partner attorney profile 3 (partner 2 / Schema type:  person)
  • Practice area / City Page 1 (personal injury lawyer in Chicago)
  • Practice area / City Page 2 (brain injury lawyer in Chicago)
  • Practice area / City Page 3 (car accident lawyer in Chicago)

Each of the pages above can include highly structured data (Schema / JSON-LD) that would give search engines, maps, and other apps rich data about granular level attributes found within the content of each page.

But let’s say “Partner attorney 1” focused on “brain injury cases”.

By associating your structured data across your website, you could in essence tie the “Partner attorney 1” profile page directly into the “brain injury lawyer in Chicago” page.

By making this association through structured data, now search engines such as Google not only has rich information and understands your “brain injury” page better, but Google also has specific instructions (via structured data) that “Partner attorney 1” is the primary attorney for “brain injury cases”.

This makes both the brain injury and partner attorney profile page more authoritative as a result of the structured data, and because of the associative properties within structured data, there is now a specific attorney tied to a specific practice area.

This same use case with structured data can be applied at scale in nearly every business type and market category.

Structured data is also a powerful way to present rich snippet content in Google search results.

Here’s an example of structured data being used to present event information in search:

Events Structured Data

Not only can structured data (Schema) be used to present additional data sets in Google, but this is also true of most search engines and apps.

Bing takes a little bit of a different approach to the way the show event data in search, but the concept is the same — structured data is being leveraged to present information in the search results:

Bing structured data

Another popular use of structured data in search results is to present reviews and star ratings along with the search results.

When you implement structured data for reviews, you can markup your page with things like

  • number of reviews
  • star ratings / average star ratings
  • person leaving the review

When you implement review markup and get stars to appear with your search result, this can help you stand out definitively in search and get a higher click thru rate.

The reason being, people’s eyes will be drawn to your search result because it’s probably the only one with stars showing.

For example, here’s the search results of the keyword phrase “forsyth county car accident lawyer“:

Review Schema in Search

Even though this website is ranked #2, it more than likely will catch people’s attention before the first search result.  And in turn, they could attract more clicks and awareness from people searching.

* Disclaimer:  Leibel Law is a client of ours

Intent of the search

So why is all this talk about Schema and structured data so important?

Because Google is trying to show search results based on a users intent, and not solely on the words typed into the search bar.

Also because Google desires to show the absolute best search results based on what’s typed into the search bar.

In order to under what the best pages are to show in the search results, Google has to have a detailed understanding of what your page is about.

And structured data helps to explain to search engines, (and maps, and apps) all of the different topics and entities associated with your page.

The more it understands the bigger picture of your page and content, the better Google will be at presenting your page for searches that are topically related to your content.

Another way to complement the structured data on your pages is to understand LSI (Latent Semantic Index).

what are LSI keyword phrases

I wrote about LSI in another post called Local SEO 2017 – It’s All About City Pages.

And here’s a quick snippet from that article:

Perhaps one of the greatest advancements in Google’s search algorithm is the move toward topical relevance. Topical relevance goes beyond specific keyword phrases, and looks to the overall relevance of a particular topic. One look at the organic rankings for Wikipedia pages will give you an idea of the power of topical relevance. Today, Google uses LSI technology to determine the overall authority of a page based on topical relevance. So instead of focusing on just one keyword phrases, Google looks at the content to see how many other “topically related” terms or phrases are used throughout the content. In summary, latent semantic indexing means that Google is trying to understand the topical relevance of your content.

Again, this is moving far beyond search results based on one single keyword phrase or just a list of 10 blue links.

The best strategy for tailoring your pages for LSI is to think more content is better.

If you have a page of content that’s only 300 words long, that’s not enough content to cover a lot of different related topics.

However, if you write a 2,500 word article or page of content, and if the content is high quality and well researched, you’ll naturally have a wide range of related topics spread throughout the content.

As mentioned in the quote above, Wikipedia is a great example of long form content that includes a lot of different related topics.

In fact, Wikipedia is so detailed with related topics, they include links to multiple Wikipedia pages throughout any given page of content.

Check out this Wikipedia page on a topics that’s been in the news lately:  Local SEO 2017 – It’s All About City Pages.

As you read through the content, notice how many additional keywords (related topics) are linked pointing to their own Wikipedia page.

In the first 2 paragraphs of that page, I counted 12 outbound links (or 12 related topics) pointing to topically related Wikipedia pages.

Now, I’m not saying the pages within your site need to be on par with Wikipedia articles!

But you can certainly put some time and effort into building out well researched content and focusing on more content per page, not less.

If you write a 2,500 word page of content and layer on the Schema / structured data, you’ll be well positioned to compete in the top search results in Google and attract traffic from a wide range of topically related keyword phrases.


The future of search will be all about voice search, AI (artificial intelligence), and machine learning that influences the search results.

This means you need to move beyond thinking of your websites and web pages as just a collection of text, images, and meta data, and start thinking about how to integrate Schema markup language to enhance your pages.

JSON-LD is the preferred markup language according to Google.

Well researched, long form content is a great way to optimize for the Latent Semantic Index.

Combined, the above strategies will help you compete in the search results for a wide range of related keyword phrases.  This is important because Google is now showing search results based on user intent and not simply the keyword phrases typed into a search bar.


Leave a comment: