Something's Different with the Algorithm

Google has updated its algorithm several times to favor the websites that perform better on mobile devices in response to this year’s increased use of Mobile Search.

“Algorithm” is the name of the system Google uses to sort through all the information it has when you enter a search. As of 2013, Google's algorithm has the name of Google Hummingbird, on account of being “precise and fast.”

Although Google actually makes minor changes to its algorithm some 500-600 times a year, “major” algorithmic updates that affect results in significant ways are only occasionally rolled out.

The last major update was the Panda 4.2 , released this July 17. Deemed more of a data refresh than a legitimate update, Panda 4.2 affected the search queries on a much lower scale than previous algorithmic updates have. Sadly, it’s too late for changes that would make Panda 4. 2 have more of an impact on queries for, once Google produces an update, changes are only applied until after the next refresh or update.

Penguin

Google Webmaster Trends Analyst, Gary Illyes, posted that the next algorithm update will be the next Penguin 4.0. Although Illyes has publicly disclosed that he cannot make any guarantees, Penguin 4.0 is expected to be real-time and continuous. 

Real-time means that the algorithm will update continuously and that there will not be specific release dates for Penguin updates after its initial release.

If real-time follows through, Google would be able to detect spam in links and would have Penguin attack it immediately. Conversely, Penguin will back-off from a link as soon as the Google’s indexer reads that the spammy links have been removed. Penguin would have the ability to re-integrate fixed links in real time, so site owners could recuperate from a Penguin penalty quickly enough. 

Penguin 4.0 will hopefully be released before the end of this year. So far, there have been 4 major algorithmic updates. Listed in descending order:

  • Panda 4.2 – July 17, 2015: More of a refresh than an update. No major changes reported. 

  • The Quality Update – May 3, 2015: After many a report of large-scale ranking charges, Google made a change impacting “quality signals.” Google didn’t reveal specifics about the nature of the signals even though their impact was broad.

  • Mobile Update AKA “Mobilegeddon”  - April 22, 2015: Google announced that mobile rankings would differ for mobile-friendly sites starting April 21st.  Impact was small.

  • Unnamed Update – February 4, 2015: Google did not confirm this update but the major flux in Google SERPs – search engine result pages—give grounds for update speculation.

In case you’re wondering, there is no stated reason to how Google names its updates. One of its algorithm updates was even coined “Caffeine.” Click here for moz’s interactive timeline of updates.

Wait… There’s More

Yes, Google announced last week that they were integrating another layer to their already complex ranking algorithm. The new signal to go into Google's algorithm is called RankBrain and is speculated to already handle around 15% of Google's search query.

What is RankBrain?

RankBrain is a machine that uses artificial intelligence to embed great amounts of written entities into vectors (mathematical entities) that computers can understand. Much like a human would, if RankBrain sees an unfamiliar phrase or a word, it can make a guess as to what in their repertoire of words and phrases might have similar meaning and provide an answer from these results.

Although not yet a replacement for the Hummingbird Algorithm, whose job is still to process searches and rankings, RankBrain has the capacity to provide never-before-seen search queries. More specifically, RankBrain is like a query processing tool with machine learning – when a computer teaches itself how to do something.

“Even though we’d like to think that a company worth 100’s of billions of dollars, like Google and Apple, should have advanced technologies that us mere mortals haven’t even heard of, the truth of the matter is, they still have a lot of manual processes and antiquated code wrapped up into their operating systems and search algorithms.”
-Forbes Oct 27, 2015

Despite its Machine Learning, RankBrain is still run and supervised by humans. For now, RankBrain’s selling point is its ability to look for the patterns that humans have programmed it to search for, and to make its decisions based on that pattern recognition. It’s still a computer.

What will RankBrain Track and Improve?

It is the hope that as billions of searches come in, RankBrain will track clicks and will start to make correlations without requiring a human to manually write it into the code.

What Does RankBrain Mean for SEO and Your Small Business?

Thankfully, RankBrain appears to only be upgrading user-experience and the approach that small business owners use for their SEO doesn't have to change yet.

Nonetheless, here are some insightful predictions:

1) RankBrain will not replace the current Hummingbird 

Not it the near future anyways, RankBrain's use of artificial intelligence only makes it fit into Google's larger system as a type of experiment.

2) More voice searches

The application of RankBrain's machine could mean that Google expects voice search to become more prevalent. Business owners should adapt their messaging for such a situation.

3) More accurate results

RankBrain's ability to learn for itself will likely make it better understand what users are searching for. RankBrain was ultimately introduced to enhance user experience and one of the ways it will do it is by helping people to better connect to information based on their search-intent. 

Businesses should be aware that more accurate searches will mean they have to keep high-quality content across all their sites and platforms and ensure their plugins are up to date. All in an attempt to make sure their page is placed high up on the search results list.

4) Customer reviews will be more important

Customers provide reviews in real language and so do people who ask a question on Google. RankBrain's ability to decipher real language will help redirect searches like "what's the best guacamole in the area for someone on a budget?" to the right location.

5) Micromoments matter

Artificial intelligence could help Google better deliver personalized search results at a specific point in time. For example, the search engine will be better able to filter results to be more relevant and trendy. 

6) Paying more on advertising could pay off

More and more users are searching through mobile. Zach Morrison, president of Elite SEM, says artificial intelligence could be the company's way of improving mobile search so that better relevancy could mean better conversions which could help justify higher ad rates. 

Your content and retention tools are as important as ever. For more resources on improving these read on.

Remember that the future of machine learning technologies will make the Google algorithm more agile and be an asset that will stay ahead in a way that no human would.

 Contact us online today with any questions or for consulting about mobile development.