We’re hot on the tails of the Intelligent Content 2015 Conference and in anticipation of upcoming related workshops. So when the question of “Intelligent Search Engines” came up on the Intelligent Content blog (hosted by the Content Marketing Institute), I decided I couldn’t fit a reply into a comment and I’ve addressed the topic here.*

We’re talking on and on about making content more intelligent these days – format-agnostic, self-describing with semantic metadata, and modular – for reuse, for omnichannel, for delivering the right content to the right user, etc. But what about search engines themselves? What happens when they become intelligent?

Search engines are already intelligent. In fact, one of the reasons that people need to get their own content more intelligent in a hurry is to keep up with the possibilities of modern SERPs. A tiny fraction of companies in the world (the ones you might see presenting case studies at the Intelligent Content Conference) are ahead of what engines are doing, but the mass market is still in the dark ages. Search engines are working bring the world forward – Google, Bing, Yandex and Yahoo! in collaboration in fact – and enticing them with the almighty “better SEO” if they step up to the challenge.

* Don’t miss the “Bonus Fun Fact” at the end.

Intelligent Search Results

Google, for example, is doing its best to add semantic categories to its page interpretation and giving tools to help communicate semantic richness from sites to the SERP. You see this all the time when you look for things and it no longer just comes up with “lists of pages” but an array of navigational and relational tools that are displaying content in fundamentally different ways than 7 or 8 years ago.

[click images to enlarge]

Results for: “da Vinci”


In this example, instead of simply returning keyword-weighted lists of pages as results, Google disambiguates between Restaurants (adapting results by cross-cutting with my location: Valencia, Spain), News articles, and general facts and background information.

Results for: “best band ever”


The image carousel returns results that are relevant, pulling key metadata (Band Names) to the surface. Even though many of the pages behind the scenes won’t necessarily actually say “best band ever” on them, these results are listed. This is much more intelligent that mapping keywords straight to pages. The engine is interpreting to get to what I wanted, based on what I said, mapping from an ambiguous concept (“best band”) to keywords, then to content.

Results for: “elton john’s kids”


The semantic categorization and mappings that are often behind the scenes can be seen right in the SERP if you search for people and note the relationships at the top: Elton>Children: Zachary (Son), Elijah (Son). This kind of relationship is being built using metadata of all kinds between all kinds of things, but as human family trees are very familiar, here it’s surfaced as a navigational tool.

Why isn’t this common knowledge?

It is and it isn’t. The technology and concepts behind this are well over a decade old, but it’s taken this long for standards and the market in general to catch up. Why intelligent content and adaptive content have taken so long to go mainstream is a whole other article.

There are several articles on the topic of semantically categorized content in search (just Google the term semantic search). There was a burst of reporting when the topic was just getting sexy in 2012-2013, especially because Apple and Microsoft were taking strides to try to close the gap, but now it’s just settling into the general background of the Search Engine wars. The bulk of communication today is (sadly) still focused on developers translating unintelligent site content to semantic structures so that Google can interpret at publish time (i.e., the very last minute).

Search engines will always focus first on the lowest common denominator. They need methods and tools every local business can adopt. Therefore, only today, when everyone, everywhere needs to get more semantically rich – on all manner of devices – did they really begin the push. And still, they’re starting with the audience who will implement the bits that they connect with directly. At the moment, that’s developers.

Content Strategists and Content Engineers must work with developers to make sure that search engines get what they want. However, developers working directly with or in small organisations can also just offer commoditized, pre-packaged services and tools (even things like WordPress Plugins).

What should you do about it?

Model adaptive content for any channel. In sessions like those at ICC, we look at modelling content to make it intelligent inside the organisation so that it’s ready for more adaptive delivery regardless of channel. One of these channels is of course the web, and therefore search engines.

Focus on user questions, not your brand. It’s the user’s journey that matters after all. In my sessions at ICC and the upcoming workshops in North America and Europe I talk about this concept. After doing a bit of extra research it turns out Google uses almost exactly the same phrase as I do:

 “When you search, you’re not just looking for a webpage. You’re looking to get answers, understand or explore.”

I say that “users aren’t looking for a page, they want an answer”. Some slides from my talk/workshop:



Mashable puts it in very similar terms as well:

“When you’re framing your content in a semantic search world, it has to be around answering the specific questions people have as it relates to that keyword. With every sentence you write, ask yourself: How does this answer the searcher’s question? You will have to focus on the natural language even if those users are still focusing on keywords.”

Rethink SEO with an emphasis on user-oriented semantics. You can look at the Moz blog and learn about conducting semantic keyword research for some insight into what this means. I especially liked this soundbite from Jon Dunn’s article on semantic SEO:

The future of SEO is about optimising your site, your content and the whole of your online presence for your customers rather than for Google, which is a fundamental shift in thinking.

Remember Search Engine providers will only focus on their own part of the value chain. 100% of what they do is focused on making sure that content can be intelligently exchanged between sites and engines. Your reuse across formats, or anything else that they don’t profit from, doesn’t makes sense for them to invest in. So if your brand is suffering because your content is not intelligent enough, no relevant tips will come up in any materials they publish. You could have great, intelligent delivery to SERPS and still be:

  • squandering budget unnecessarily or missing opportunities during content creation, translation, governance
  • not adapting sufficiently for user needs or failing to optimize content personalization
  • underperforming on alternate channels
  • allowing inconsistent messaging, facts and branding to get out to market.

That’s why events like the ICC are so important and why I’m running my legs off trying to raise awareness beyond the web and search engines. The more educated the market is on the omnichannel benefits of intelligent and adaptive content, the more fun my job is.

Define and manage your own intelligent content semantics.  Your organization’s defined semantics can then map to public standards like Schema.org on the way out to the web. This keeps the brand in control and makes sure that content is about your users and business goals across all formats, not just about what Google and SEO wants. These same concepts apply to app design, content reuse for print and packing material, social, emailers, intranets, and on and on. Never lock your content into any single channel.


Structure your content in the way that makes sense for your users and business. You can adopt several things from public formats into your internal standard, but your content should be yours. That way:

  • You’re always future-proofed in case this or that consortium to get overtaken (e.g. something called “Microformats” were recently supplanted by Schema.org before many people even heard about Microformats)
  • You’re not duplicating effort for your non-web channels.

I hope this cleared up a few things and I hope to see you at a workshop or conference soon!

Further reading

(Techie articles are labelled as such)

Bonus Fun Fact

Although it’s not “Intelligent Content” at all, I can’t resist mentioning this if I’m talking about what an “Intelligent Search Engine” looks like: Google is starting to provide intelligent interpretation of non-textual content assets as well. If you, or anyone you know, has their photos backing up to Google Plus, try searching them with any of these keywords:

  • Dog
  • Face
  • Arm
  • Sky
  • Beach
  • Food
  • Tree

The results are frankly breathtaking. This example for “food” is turning up results that have not been tagged in any way whatsoever.


These images are straight from the camera to my personal cloud space. The files don’t even have names, they have only auto-generated identifiers. The engine is looking at the image and interpreting it just like a human would.

In true Google style, as the results are still imperfect, they’re not even telling people about this jaw-dropping bit of intelligence. Imperfect is sometimes amusing: my search for “Dog” turned up several dogs, a fish, two ducks, a group of kangaroos and a baby (note, he wasn’t facing the camera).

My food example has two results that are from the front and inside of a restaurant, and even I refuse to completely believe it managed to make that connection… although maybe with GPS + the restaurants location data + the fact I took some food pics inside? It makes one quiver to imagine…

Try it. Amuse your friends. Let me know in the comments if you find any other keywords that work. I found this list by random experimentation.