The Truth Behind Hummingbird: Google’s New Old Car

Many of us expect a lot from the Internet, and for the most part, it’s do-able. However, Google’s recent announcement of the new Hummingbird search algorithm triggered guesses of functionality online that bordered on something Isaac Asimov would dream up.

Here’s the truth: Google hasn’t overhauled their search algorithm in about three years, instead opting to put out updates that address certain bugs or gaps that they’ve found. Hummingbird is new, and it is meant to get better results for conversational search, such as “Where can I find the best steak in New York City?” versus before, Google would be more prone to understand something in caveman-speak: “Where best steak New York.” Note that I say more prone, because Google wouldn’t have broken down prior to Hummingbird had you actually asked your question, and your results could have been the same as they are with Hummingbird now.

What this doesn’t mean is that Google has created a search engine that understands conversation, colloquialism, or relative notions of “best,” “worst,” etc. If Google were to do that, they would probably be supreme masters of the universe by now. (We’ll give them a few more years to get there.)

The truth of the matter is that algorithms haven’t evolved since the 70s, and to build a search algorithm that could actually understand queries and give you options based on what you’re asking would really be creating the first sentient machine. Google, like all of us mere mortals, is limited by the software capabilities that currently exist.

Make no mistake: leaps and bounds are being made all the time. It just happens to be the case that those leaps and bounds have a glass ceiling no one has broken yet. While hardware advances to record-levels of processing speed and data storage, the software that can actually decipher, categorize and turn all those tidbits into something we can understand is—so far—the stuff of science fiction.

So what changes with Hummingbird, then? Google’s new search algorithm does represent their devotion to working toward bringing software to a human level, versus relying on people to dumb-down to machine level in order to be understood by a computer. Hummingbird also has the potential to make real conversations, such as forums, comments, and reviews much more valuable to brands for search engine optimization purposes. While page rankings are still many brands’ holy grail, Google is showing that keyword optimization through content management systems (like WordPress) is taking a back seat to all of the human interaction that actually proves your brand is front-of-mind with consumers.

Sorry, Internet. Hummingbird doesn’t understand exactly what you mean by “most awesome video” or “cutest puppy” yet. Brands should take heed of this move, though, and understand that search engines are moving to a more human-centric online experience, and optimizing your stuff online is a matter of real people performing real actions.

In a move to minimize technical “tricks” that win page rankings, Google’s Hummingbird is, at most, a very, very, very small step toward a sentient system that actually understands questions.

Ultimately, it is currently closer to Roomba than to iRobot.