Content Strategy and Natural Language Search

I was sitting at my desk looking at the Starbucks logo on a cup of coffee I’d just purchased. I was curious. What was the story behind that logo, anyway? I Googled: “what is the starbucks logo meaning?”. What I was expecting was a link to another page, but what Google provided right on the search result page was even better:

In a search for a way to capture the seafaring history of coffee and Seattle’s strong seaport roots, there was a lot of poring over old marine books going on. Suddenly, there she was: a 16th century Norse woodcut of a twin-tailed mermaid, or Siren.

Google managed to answer the question almost perfectly. What was even better was that Starbucks got me really interested in hearing the rest of the story. I was curious now, how well did Google answer questions like these? I Googled: “what is the nike logo meaning?”. Again a quote appeared trying to answer the question, but this time the answer wasn’t quite right:

The swoosh is the logo of the athletic shoe and clothing manufacturer Nike. It is one of the most recognized brand logos in the world. The Nike “Swoosh” is a corporate trademark created in 1971 by Carolyn Davidson, while she was a graphic design student at Portland State University.

The big difference from the two logo search results, from what I can see, was the source of result. Starbucks wrote their own story on their website, but the Nike result returned a Wikipedia page.

When Google (and others) really master natural language queries, these are the details that will be important. I continued my experiment and searched for “what is nike fuel?”, and got the result:

NikeFuel is a proprietary unit of measurement for tracking fitness activity developed by the athletics company Nike, Inc. The exact formula for the measuring unit is proprietary and depends on the device or service tracking it.

A pretty boring, clinical explanation. What’s interesting is that is actually the first regular search result here, but it’s not been used in the Google result page (which might be the only thing a customer reads). I’ll hazard a guess, and say that the content was just too poor to be considered for the natural language result.

In SEO, we are used to writing content to make search results, but soon we’ll see an expansion of that. Instead of writing for keyword density, we are going to have to write our content for these natural language systems. The good content authors will be the story tellers, who can tell interesting stories for us, but also speak plainly enough for the bots.