Artificial Augmentations: Adapting SEO Strategies with the Rise of A.I.

Artificial Augmentations Adapting SEO Strategies with the Rise of A.I.
Artificial Augmentations: Adapting SEO Strategies with the Rise of A.I.

Back in 2015, the organization behind Formula E, basically a fully electric version of Formula 1, announced Roborace, a motorsport championship featuring electric-powered single-seaters with full autonomous driving capabilities that will run as a support series for the 2016-2017 Formula E season. Long story short, it didn’t happen but they are still working on getting the inaugural race off the ground. In November of 2017, a time trial competition was held in Paris between the development car and Formula E reporter Nikki Shields. Shields posted a time of 1:26.6 while the Devbot ran a lap of 1:34.0. Not quick enough to be competitive but with the pace of development A.I. is currently on, it won’t be surprising to see Devbot running a quicker time than an amateur driver by the end of this year.

While no commercially available vehicles are fully autonomous as of this moment, Alphabet’s Waymo is currently testing its car in public traffic and Singapore-based nuTonomy opened up its robo-taxi service to select members of the public. The above examples are just illustrations of the more obvious ways A.I. is about to transform the landscape we’re currently living in. One lesser-known change A.I. is bringing but also directly related to our day-to-day activities is how it’s going to change the internet, specifically how search engine works. In October 2015, Google first confirmed the use of RankBrain, a machine learning system that helps Google’s own search team in processing search queries. In fact, Google went as far as announcing that RankBrain is the third most important criteria in its ranking factor, behind links and content, forcing SEO services to figure out just how RankBrain works and how best to adapt to this implementation of A.I.

Data to A.I. is what brains are to zombies

Machine learning is exactly that, it is an implementation of A.I. in which the machine ‘learns’ from data it is being fed to. If you’ve ever used Netflix, you should notice how the streaming service gives you user-specific recommendations on its ‘Top Picks for …’ section. That’s Netflix’s machine learning A.I. doing its job where it suggests films and series for you based on your and millions of other users’ watching habits. In Google’s case, RankBrain is mainly used to process queries by putting the exact words in context and figuring out how a previously unseen search query relates to other queries that the algorithm has processed previously. This is in essence an extension of Google’s Knowledge Graph, in which the search engine switches its nature from processions strings of text into trying to figure out what that string means and presenting you with what it thinks is the proper answer.

It sounds a little bit complicated but here’s an illustration to help you. For example, when I put in the words ‘film with talking beetle’ into Google, the results that pop up are about Herbie, the 60’s original film The Love Bug and the 2005 revival with Lindsay Lohan. You see, the algorithm understands that when I typed beetle, I was talking about the iconic Volkswagen and not the actual bug species and even though Herbie never actually talked in the films, Google understands that I was talking about an anthropomorphized car. Now, this is actually a pretty welcome change for the users but how exactly is this going to change for the other party, i.e. the content providers themselves and how can they adapt to this? It’s not exactly a fundamental change, but there are some measures you could take to optimize for this:

  • Optimize your keywords

As illustrated above, while searching for films on Herbie, I used a query that is only tangentially and somewhat ambiguously related to the iconic car but the result that popped up still matches with what I was looking for. This is also true when you’re creating your contents. Instead of focusing contents on a single keyword or phrase, try to think of contextually similar keywords, words and queries that might sound different from your targeted keywords but ones where the intended meaning is in line with what you’re looking for and find an exemplar that could theoretically represents those keywords you’ve collected. There are quite a number of online tool you could use, free or otherwise, in which you can find related keywords and one that is getting the most search hits.

  • Optimize for people instead of machines

Sounds a bit contradictory given how RankBrain is a machine learning system but RankBrain feeds on data given by actual human users and try to interpret them to give the most humanly possible response it could muster. Again with the Herbie example above, anyone with knowledge of the iconic Beetle would know right away I was talking about Herbie but a machine might not. Try putting the example query above in other less-sophisticated search engines and note the difference. One of the easiest way is to use a conversational tone, but not infested with slang and chatspeak. The tone should be somewhere around a thesis you wrote for college and your chat history with a close friend.

  • Consider user-generated contents

To be more specific, you should always be mindful of user reviews. Google mines data from almost everything, for example say you run a curry house specializing in Japanese-style brown curry. One of your customers posted a review on Google claiming that your food is great and with a price to match. Now, whenever someone in your area search for ‘best curry’, your place might come on top but adding a modifier, ‘best budget curry’ like so, might leave you out of the results entirely. This, by the way, isn’t a a bad thing as depending on what your intended market is, it could even work entirely in your favor. One other thing worth mentioning is that it is possible for customers to ask questions or leave comments in your Google My Business (GMB) listing and it is in your best interest to monitor this.

  • Make use of data

Tools such as Google Analytics are there for your own benefit, use them. Machine learning systems like RankBrain gets smarter and smarter by consuming data and even though it doesn’t work that way for us fleshy creatures you could still use those same sets of data to gain new insights on your customers. Truthfully though, stats and charts aren’t exactly for everyone and if you’re the type who’s just not cut out for this, find an SEO services that do.

A.I., machine learning and autonomous vehicles, those phrases used to be simple buzzwords decades ago. Things that science fiction literature and films have hinted on but always seem out of reach when it comes to reality. Well, not anymore. The rate of technological progress in the past ten years and the seemingly overnight democratization of the internet has opened up the world and the massive amounts of data it processes each day to anyone who knows where to look. Back in 2015, experts predicted that an A.I. would be better at Go than humans by 2027, in March the next year, DeepMind’s AlphaGo played against Lee Sedol, one of the best Go player at the time in a 5-game match. The final tally was 4-1 for AlphaGo.

RankBrain doesn’t bring with it fundamental changes that revolutionize search engines as we know it, but it does make them a whole lot smarter and unlike other algorithm changes, the nature of machine learning itself means that it’s going to keep changing. Things that we consider to be best practices right now might not be the same in the next 5 years. SEO itself has always been a changing and thanks to Google’s lack of transparency, murky landscape. Now that A.I. is in the fold, that landscape is just going to get murkier.