expertSEM

Wired Magazine’s exclusive article on the Google Algorithm is one of the best explanations I’ve ever read of how Google does what it does to pursue its mission of organizing the world’s information and making it universally accessible and useful. (If you do paid search marketing or provide search engine optimization services and you haven’t read it yet, shame on you.)

The article explains – in clear everyday language – how Google’s engineers use their own search data to update the technology to recognize complex language patterns, like synonyms and multi-word phrases which mean different things than the individual word parts. Here is an excerpt:

Take, for instance, the way Google’s engine learns which words are synonyms. “We discovered a nifty thing very early on,” Singhal says. “People change words in their queries. So someone would say, ‘pictures of dogs,’ and then they’d say, ‘pictures of puppies.’ So that told us that maybe ‘dogs’ and ‘puppies’ were interchangeable. We also learned that when you boil water, it’s hot water. We were relearning semantics from humans, and that was a great advance.”

But there were obstacles. Google’s synonym system understood that a dog was similar to a puppy and that boiling water was hot. But it also concluded that a hot dog was the same as a boiling puppy. The problem was fixed in late 2002 by a breakthrough based on philosopher Ludwig Wittgenstein’s theories about how words are defined by context. As Google crawled and archived billions of documents and Web pages, it analyzed what words were close to each other. “Hot dog” would be found in searches that also contained “bread” and “mustard” and “baseball games” — not poached pooches. That helped the algorithm understand what “hot dog” — and millions of other terms — meant. “Today, if you type ‘Gandhi bio,’ we know that bio means biography,” Singhal says. “And if you type ‘bio warfare,’ it means biological.”

Fascinating. If you haven’t read the article yet, read the entire thing right now and then bookmark it to read again later.

I’m even more personally interested right now because I’m experiencing a minor Google matching fiasco with my domestic airline campaign. My client’s brand is properly trademarked and some new ads I’m testing kept getting disapproved – for trademark reasons. Thankfully, my Google agency team is helping with an interim solution. But what we also discovered is that this is not in fact an issue with my clients own brand terms. It’s because the call to action in my new geo-targeted copy reads, for example, ‘Get our lowest fares from Philly!’. [insert thoughtful pause] Get it? Like the cream cheese! My ads were being disapproved because of the cream cheese trademark, not the airline trademarked terms.

Curious, I did a one-keyword search for ‘Philly’, wondering if the Kraft cream cheese brand is stronger than the city when it comes to SERP results. It is not. In fact it doesn’t appear up to page 5 of the organic listings and whomever is doing PPC for the cream cheese is not bidding on that single term. The Google related searches feature shows a link for ‘philly cream cheese’, but that feature is at the bottom of the page now or you have to click the plus mark for ‘Show options’ at the top to get there. So my little test proved the city takes precedence over the cream cheese in SERP results, but clearly the opposite is true with the trademark in my ads.

The end of my story is that I’m still fascinated, my ads are finally live, we’ll never really know everything there is to know about Google’s algorithm, and maybe the Google engineers and the trademark team need to go to happy hour together?