Today a new innovative product from the British data science specialist MathSight, is lifting the lid on the secret ingredients to Google’s Panda algorithm.
Google is continually updating their software to combat unethical search engine optimisation (SEO) techniques and currently uses a number of algorithms including Google Panda. Panda was most recently updated by Google on the 25th September. This software is designed to stop sites with poor quality content working their way into Google’s top search results; however these software updates can also have a detrimental effect on legitimate business dependant on web traffic to survive in extremely competitive markets.
Germany’s justice minister Heiko Maas recently called on the search company to reveal how it creates search engine rankings, telling the Financial Times that Google had to become more “transparent” about its highly secret algorithm that underpins its search engine rankings.
Google’s is dominating internet search across Europe with a market share of around 95% and has been accused of exploiting its dominant position to promote in-house services such as shopping, flights and restaurants.
Mr Maas said: “We know, because we all use the internet, that whoever isn’t at the top of the search results virtually doesn’t exist”. It is inconceivable that Google will ever disclose the details of the algorithm to regulators and competitors, since such transparency would make its search engine a target of spam and expose the business secrets that give it competitive edge. This is why businesses need to seek an alternative way to deconstruct the Panda 4 algorithm and its future updates.
It is reported that the recent panda update lost certain businesses up to 40% in web traffic and until now, businesses have had to rely on inaccurate SEO techniques to determine how their website search rankings are rewarded or punished against the Google algorithm updates.
MathSight has used sophisticated mathematics to build an artificial intelligence platform that predictively models a number of search engine algorithms to identify exactly which aspects of content, web design, and site architecture, along with their inter-relationships, traffic data and other key performance indicators determine the cause of change to search engine traffic. Businesses are then able use this intelligence, to implement the exact changes to increase their search engine ranking.
The results were conclusive and unveiled specific key findings that would reward or punish a website following the Google update; such as the proximity of body copy to the top of the page and the exact ratio of rare and common words to the total body copy. Meaningful article Meta titles were also heavily rewarded.
MathSight used Bayesian mathematics, a set of mathematical rules used against new data to continuously update beliefs or existing knowledge. This enabled segmentation of vast amounts of data, which in turn made it possible to decompose the content into their various attributes, identify specific patterns and test emerging patterns for cause and effect on traffic using variance analysis.
The same techniques were applied by MathSight last year to reverse engineer Google’s Penguin algorithm where it was revealed that Google were examining the readability of content to decide whether a site’s traffic sources were genuine public relations coverage, or fabricated by their SEO provider.
MathSight is the brainchild of search engine optimisation specialist Andreas Voniatis, a trained and qualified management accountant who switched careers in 2003 to become a SEO specialist.
After holding various Head of Search roles for award winning agencies and prestigious startups, Andreas identified a need for empirical data analysis in SEO, which would enable him to dissect the inaccessible Google algorithm
Following a chance meeting with Frank Kelly an experienced Systems Architect with a Master of Engineering from Cambridge under his belt they founded MathSight.
Andreas said “We are excited to announce to the world the some of the unique findings from our highly confidential research, which enabled us to develop the MathSight platform. Nobody else in the industry appears to be addressing the huge collection of data with advanced mathematics, which is the only way to unlock its valuable potential and keep one step ahead of the search engines.
“We look forward to integrating with more bespoke enterprise level, SEO tools to ensure business get the best possible chance of visibility across search”.