Solutions
Integrations

Integration Spotlight: How to use Amazon Comprehend to power better SEO results

Natural language processing, or NLP, has been making news for years related to its growing impact on SEO and its adoption by Google and Bing. This significance only grows as people get more and more comfortable with using voice-to-text features on their phone, handhelds or even desktops.

March 23, 2021
Search engines like Google and Bing are working to deliver search results that respond more accurately to questions phrased in conversational language. Now and into the future—especially with the growth in voice-activated search—meeting complex requests phrased in our natural language will be key to maintaining optimal search performance in the future.
2 Min Read

Brightspot’s integration with Amazon Comprehend brings the power of natural language processing directly to your publishing experience. When used within the SEO tool set of Brightspot, Comprehend can now identify the main entities of your content and provide their salience score. Salience is the relative significance of each analyzed keyword to the overall content. The data provided by the tool can give us a glimpse of how a search engine like Google or Bing might interpret our content and further our SEO understanding.

A real world example of how Amazon Comprehend can inform and influence SEO

Today’s content management system marketplace is highly competitive with multitudes of offerings flooding the space. This means the SEO competition for all things related to CMS platforms is equally competitive and unforgiving.

In this article we will show you how the use of Amazon Comprehend’s Natural Language API allowed us to identify SEO keyword issues within our own content plan.

Great content is the best SEO

A few months ago we wrote an explainer article, "What is a content management system?"

Along with the ultimate goal to educate readers about the options and the CMS space, we had written the piece to target "content management system" as the primary keyword term.

Brightspot’s integration with SEMrush told us that this keyword had a high search volume. A deeper look at the SEMrush data also told us that "content management system" would not be easy to rank for, with a keyword difficulty score of 79% bordering on "very hard" (considered the most competitive terms requiring excellent on-page SEO, link building and content promotion efforts).

SEMrush keyword ranking details for content management system
SEMrush keyword ranking details for "content management system"

This is obviously somewhat of a generic topic but one which Brightspot is certainly qualified to address. Our content staff did their SEO due diligence and researched what the current top-ranking results were and then wrote our own unique content following the golden rule of SEO—content is king.

Once we had our new article and felt that it not only targeted the keyword "content management system" but added real value to any user searching on that term, we published and waited. And waited. And waited some more. Our finely crafted content never broke Google’s top 100 search results. We were able to validate that the content was being indexed. It just wasn’t ranking.

Amazon Comprehend to save the day

Frustrated with the lack of results on our SEO content efforts we decided to take a deeper look.

A snapshot SEO audit for the piece told us that the:

  • Content is unique;
  • Content contains the targeted keyword multiple times;
  • Keyword is in the title;
  • Keyword is in the headlines;
  • Content is linked to internally;
  • Crawl depth is <= 2;
  • Images are relevant and have good alt text. 

On the surface it looked like we had checked all of the boxes for SEO best practices. It's at this point Brightspot’s integration with Amazon Comprehend saved the day.
When looking at the results from Amazon Comprehend we could see that the natural language API came back with some surprising results. The main entities of the page were not "content management system" as we had intended.

It turns out that our content contained multiple variations of the target keyword and this ultimately diluted the target keyword's significance. So while "content management system", "CMS”, "CMS’s", "content management systems", "content management platform", etc. are all contextually relevant, they are not the same.

The other thing that stood out was the low salience scores of the returned entities. This was another indicator that our targeted content wasn’t as focused as we had intended.

Output from Brightspot's Amazon Comprehend Natural Language SEO feature
Output from Brightspot's Amazon Comprehend Natural Language SEO feature showing diluted salience scores for the target keyword.

Armed with new knowledge and the tools to make informed changes we rewrote the article and used Brightspot’s integration with Amazon Comprehend to validate that our target keyword was indeed now a relevant entity within our content.

Brightspot Amazon natural language api for SEO
Amazon Comprehend Natural Language scores reflecting higher salience scores following initial analysis.

As of this writing the article is now climbing the rankings for organic search and improving week over week. A measurable impact made possible with data insights gleaned directly from Brightspot’s integration with Amazon Comprehend.

To learn more about Brightspot's integrations with Amazon Comprehend and other partners, contact us today to get a demo of our award-winning Content Business Platform!

For more details on Amazon Comprehend and Brightspot

Amazon Comprehend uses natural language processing (NLP) to extract insights about the content of documents. Amazon Comprehend develops insights by recognizing the entities, key phrases, language, sentiments and other common elements in a document.
Amazon Comprehend suggests relevant tags, so your editors no longer have to spend time searching and searching for relevant options. See how it works.
Choosing the right tags can be time consuming and lead to missed opportunities. Here’s how Brightspot solves that problem with Amazon Comprehend.

Share
Miles De Feyter
About the Author
Miles brings expert front end development and management expertise gained from his 20+ years of leading web dev teams, including design and development, monetization, and tactical operations. Prior to Brightspot, Miles was a development lead at AOL, a contributing member of the W3C who helped the HTML Working Group define what would become the HTML5 recommendations, and led AOL News' effort to implement and help define the rNews microformat which would later become today's standard for NewsArticle schema.