Your Essential Guide to Writing Content for Semantic Search

Is your content mobile-ready?
If you’re anything like me, when you first heard of “semantic search,” “semantic web,” and “entities,” you thought – not more monsters to tame! When did they think these up?

Actually, they are all part of the same “monster” and they’ve been around a while, lurking under the bed waiting to come out of their dark corners. But now that they’re out and functioning, they’re more like the characters in “Monsters Inc.” than the freak in “Frankenstein’s Monster.” Get to know them and you may actually like them.

True semantics

Here’s a simple way to think of them:

  • Entities are attributes like names, numbers, addresses, or other items of data that reside on the Web.
  • Semantic Web is where the entities reside and how they’re coded to make them accessible to anyone from anywhere.
  • Semantic search is the relation between all the entities that give wider meaning to our content and the searches for it.

Behind that simplicity is a more complex explanation and you can find it by Googling any of the terms.

What will be more useful to you as a content producer is how you use them to build better content and attract more traffic.

Hummingbird is semantic search in flight

Google Hummingbird is semantic search in practice. Bing has its version, too, although they haven’t given it a name. You can see it in use in the way Google and Bing display information to answer your queries with sidebars (Knowledge Graphs, Rich Snippets) or carousels of information you didn’t ask for but that help you learn if your search engine understood your query.

Do a search on a topic and a Knowledge Graph will display data on locations or people or organizations associated with, conferences or speeches about it, with links.

Website programmers use semantic coding (RFD, RSS, OWL sound familiar?) to highlight data in text documents not yet structure for accessibility and, thus, create structured data to populate the search results to provide immediate answers. They also use this semantic language to create structured data in other data formats using predetermined codes provided at that allow for standardized use across the Web. The creation of all this “semanticized” data and the engines to access and determine its relationships is the actual Semantic Web.

So how do you make use of semantics?

Nine ways to tame the semantic beast

#1. First, don’t be afraid of it

It isn’t really a monster, it’s more like a robot you program to make life better. Robots are good, right?

#2. Write with Hummingbird in mind

Remember what we said about writing for Hummingbird?

Google’s new search algorithm (September 2013) is all about building nuance for your keywords:

  • Move beyond the single keyword model and work with long-tail keyword strings, incorporate keyword synonyms, and add links to quality pages whose association also add meaning to your content.
  • With entities or data mined from trusted sources on the Web or social media, you can take it a step further, finding ways to link to data to build even more nuance.
  • Create content that answers the questions search users are asking when they do a search!

#3. Broaden your keyword reach

Plan out future content for building breadth and nuance.

  • Create a list of keyword variations to produce content about related topics. Searching those keywords, you may also find related authoritative content for linking.
  • Extend the list to include theme or concept related words or phrases. Example: If your topic is cars, a related theme might be mini-vans or SUVs; if your topic is lawn care, a related concept might be lawn sprinkling or pool care.
  • Look for keywords that anticipate your readers’ next questions or needs. This is a lot like cross-marketing, only the effect is extending meaning across pages as well as leading the reader to the next logical step in your website.

#4. Get your programmers involved

Your programmers should know more about the semantic Web and how to code for it.

  • Make available what data you can that both builds your authority and creates nuance for your brand.
  • You want entity extraction (pull data from existing unstructured Web documents) and enable new data with schema coding.

#5. Build site authority

  • Make your pages informative, feature unique information or your unique take on existing information, link to additional data, share via social media engagement, and become a source – first to report rather than being a second or third voice.
  • Be the best at presenting information. Make quality, depth, and authenticity your hallmark. (See “Authority – How to Build it into Your Site.”)

#6. Enhance site quality

  • Provide value.
  • Write well, ensure accuracy, never duplicate articles, and avoid spam at all costs.
  • Find ways to package data that others can access and use the semantic web to pass it along.

#7. Create volume

  • By this I mean create both depth (long material) and breadth (lots of articles). This creates plenty of opportunity for the search engines to understand the nuance of your site and your content.
  • The more you link between related content, the better.
  • The faster you do it, the more quickly you can benefit from the indexed relationships.

#8. Engage through social media

  • Build relevancy through shares, likes, RTs, and +’s on social media. More than ever, you need an active presence on Facebook, Twitter, LinkedIn, and Google+.
  • This isn’t just about networking, this is about showing that your content has authority and quality and all this interaction establishes nuance for your content.

#9. Establish Authorship

Finally, although some have pooh-poohed Google authorship, there is some evidence that establishing authorship creates authority and builds nuance and veracity for the topics you cover.

  • It doesn’t take much to set up and I can’t think of a negative to doing it.
  • If you already have a Google account, why not?

No monsters here

Really, most of the work in semantic search you may already have been pursuing because of Google Hummingbird. The other half requires a little understanding but is best handled by programmers and then the usual link research. So making use of entities through the semantic Web to enhance traffic by way of performance in a semantic search isn’t as monstrous as you may have thought – right?

Bien écrire!

Your Turn

What are your thoughts? We’d love to hear from you. Please share your thoughts in the comments below. Thanks.


  1. Alan Eggleston says

    As an addendum to point #9 and authorship, I just read an article on Search Engine Land that Google admitted at a Search Marketing conference that they do indeed use author rank, especially for in-depth articles, “for authors with high-ranking authority.”

  2. says

    Thanks for a great explanation of semantic search, Alan. I liked your steps to working with semantic search. There’s an interesting whiteboard Friday just out by Rand Fishkin in which he explains that it’s still important to know keywords. Even though keyword density is no longer important, having your keywords sprinkled in the right places still is. As are keyword variations, as you mentioned. And of-course authorship and authority. Interesting times ahead.

    • Alan Eggleston says

      I’m totally on board about knowing and using keywords. We just need to extend them and be aware of how differently people are using search and how differently search engines are feeding answers to searchers. Knowing that and how to feed the right information to them is the key. Understanding mobile and voice search is vital to future growth.