Google Ranking Factors
Based on Experimentation, Applied Logic & a Whole Lot Of Guesswork
2016 – 2017
2016 – 2017
This website was created in my personal time. Why not buy me a coffee?
(NZ $5 secure payment via Paypal)
Thanks for taking the time to engage in the content in this section. Because Google doesn’t tell users explicitly what goes into assessing rank, people like me have at least a couple of choices around discovering what the factors might be. Usually, it comes down to either of these two things: 1) We listen to what other people have discovered through testing, or from general observation. This is the main way in which I learned SEO myself, by reading a lot of articles and learning from others. 2) We experiment on our own websites, but that presents a few hurdles, like how we can isolate cause and effect. Or, how we can experiment on a single factor at a time, which would take many years to complete, and after which the algorithm will have changed and we’d have to start again anyway.
Reality means there is a blend of the above two factors, plus an element of guesswork, logic and assumption. As such, my list of ranking factors would have to be labeled as a work of fiction. I don’t really know what the ranking factors are, but logic and reasoning has led me to believe in a fairly extensive set of parameters. If you think that I’ve missed something, please let me know via the suggestion form. If you are learning SEO from scratch, please enjoy reading, and maybe testing my ideas for yourself. Good luck!
Scale: While some ranking factors might have a simple plus or minus factor applied, like with the mobile friendly factor, rank adjustments are often scaled relative to how far from the ideal position they might be. I’m just making guesses here as to how strongly the positive or negative effect may be applied. Nobody really knows for sure.
Boolean: the factor is applied in a simple yes or no condition and isn’t typically scaled in any way. A perfect example of this might be “is or isn’t mobile friendly”, or maybe “does or doesn’t use https”. Tip – the mobile friendly consideration is applied on a per-page basis, not sitewide. You can have non-mobile-friendly pages in your website and still pass the mobile friendly test on other pages.
NO | YES
Bell Curve: the factor is applied if the website does not behave as the expected norm where all sites in total, or per genre are plotted against each other. The bell curve is a statistical model whereby the majority are usually grouped closely together (usually near the centre – but not always), while at extremities in range there are ‘outliers’. Much like naughty children, any site that does not behave like the average is called a ‘deviant’. Outliers and deviants typically raise flags and may get positive or negative attention depending on how far from the norm they lie – again, like children. The total volume under the line of the graph represents 100% of the sample, and at the extremities, rate of occurrence is more rare. In the model I would use, there’s no maximum on the x axis. For example, websites in the genre “plumbing services” would probably have fewer backlinks than websites in the genre “fashion news & events”, simply because of the nature of the topic. That means the ‘normal’ backlink profile for a site from each genre compared to each other would be very different. That likelihood places the peak of the curve at different places on the graph. For the plumber’s website – it would be closer to zero. But for the fashion news website, it would be somewhere out to the right of the graph, where x is much greater. So a fashion news website with only as many backlinks as the average plumber’s website might be affected differently by the algorithm than the plumber’s site is, simply because it lies below the norm for its genre. Each genre will have its own statistical model applied and assessed for deviation.
This all means that performing SEO on different genres requires slightly different performance variables to be satisfied, and those variables can only be calculated against the other websites that Google has classified into the same genre. How do you know what genre your website fits into? I hope you can figure that out yourself, but sometimes it’s possible to be classified differently to what you might hope or expect. Here’s a clue: Check your content keywords list in Google Search Console. If you can show the first 20 words to anyone and they understand the theme of the website very clearly from that, then you might have nailed it. But if the theme of the website seems a bit mixed or confused based on those first 20 words, then you might have missed getting into the right genre classification and your site may well be rendering for less-than-relevant search phrases in organic search.
By the way, by my reckoning, genre classification works on a statistical model called ‘Set Theory’ and is applied through semantics. Set Theory is popularly typified by two overlapping circles where common features are listed in the overlapping section.
Geometric: the factor is applied progressively more or less across the site or page depending on the intensity of the issue. The zero point (the point where no rank gain or loss is applied) may not necessarily align with zero occurrence of a negative or positive feature. The curve may also be double-ended – meaning, it may arc upward or downward after the condition line crosses y=0, or may simply diminish to zero. I’ve forgotten the maths behind these curves, but I vaguely remember equations like y=x²-2 creating a ‘u’-shaped curve that crosses the y axis at -2. Don’t ask me more than that. I’m a linguist, not a mathematician.
Arithmetic: a factor that is applied with a simple ratio based on volume of negative or positive features. This forms a straight-line relationship between the feature and the effect. The reality is that it’s typically more complex than this, but I’m referring to a sort of basic underlying pattern of cause and effect. The steepness of the line indicates how strongly a factor is applied. The point where the line crosses the y=0 mark is where the condition has no effect, good or bad. The condition may exist naturally to some degree, in many websites as the norm. That means the zero point would represent the norm, and not necessarily “perfection”.
A reminder: The above factors are a result of a mixture of things. Some have been tested, others are just pure guesswork. Proof is not that easy to establish because usually there are so many different elements that change in a website when SEO services are applied that it’s not possible to state that a gain or loss of rank is related to anything specific. Use this list solely as a guide to my thoughts and opinions on the matter, so treat it strictly as a work of fiction.
Beware of any SEO services provider who claims to have a strategic partnership or ‘inside relationship’ with Google. Being a “Google Partner” (which I am) will provide no benefit or preferential treatment in Google Organic Rank in any way. Google will not confirm exact details to anyone about individual elements in the algorithm but will give general quality guidelines. All SEO services that I deliver are regarded as ‘White Hat’ SEO and will serve my clients well into the future because I adhere to those guidelines. I do not have any direct affiliation with Google.
I do not provide any SEO services to clients direct. Sorry! If you would like me to connect you with a trusted provider, please fill in the form here and I will pass on the details to one or two companies I trust and know will do a great job for you.