T2:Judgement Day is here, well not quite, but the machines are learning figuratively speaking from a Search Engine point of view.
Stanford University in California describes Machine Learning as “Machine learning is the science of getting computers to act without being explicitly programmed.”
Back at a Google Partners event I attended last September in Dublin, Google placed a lot of emphasis talking around Machine Learning and how this was going to be the future of search.
Not just paid search but search as a whole, and this now appears to be case.
Google, it appears, are pushing to evolve it’s machine learning in order to provide it’s users with the most relevant information/pages/products/business and to ultimately provide more satisfied users.
The search rankings will then be more defined by a ‘satisfied user’ rather than historical black hat attempts to manipulate the rankings. This is just another step away from the ‘old’ tactics.
Of course this makes the role of SEO’s, myself included, slightly more taxing to really understand the metrics which are involved in creating a ‘satisfied user’ outcome.
Webmasters this year, more so than previous years, have been seeing their own sites and their clients, rankings fluctuate on a regular basis.
Some have put this down to Google’s specified updates (including mobile and the current speed update) but the fluctuations are more regular than they have ever been.
This then ties in with the learning ethos – test, learn and roll out, with the occasional revision/roll back if required and then of course repeat.
Any strategies which are pulled together have to be more fluid and flexible than ever before, as to ensure performance is still maintained, whilst these machines are learning.
Clearly this isn’t new to the SEO world, as Rankbrain was Google’s the first recognised Machine Learning technology and was rolled out back in 2016.
This saw fluctuations across the board for large scale sites and had previously safe keyword performance affected by this learning. But now different sites were providing better metrics back to Google for them to make the decision on whether or not they had served the right content.
We knew that the quality of the site would be just as important as the link acquisition efforts but the machines will be looking at all areas of a site.
Such as mobile (rolled out March 2018), speed (Being rolled out now July 2018) as well as great engaging knowledgeable on page content and the natural link placements.
In my own experience knowledge based long tail terms have become a lot more valuable than short exact match terms.
So what do SEO’s need to do?
Well the new(ish) world which we are embarking on is evolving and regular understanding and in-depth analysis is required on a more frequent basis. It now cannot be down to just sticking to tried and tested methods but to look at what is driving this ‘satisfied user’ experience and how your site or your client’s site meets the ‘satisfied user’.
That’s obviously the million pound question for SEO’s.
The same technical aspects will be in place so ensuring there are no glaringly obvious technical errors is always an easy starting point – let the Google Bots on the site and direct them around through a well-constructed site, with the correct content and markup on the page.
As well as ensuring the sitemaps are all correct and in place, this just makes it easier for the machines to understand a site.
The machines are getting smarter in finding the right users, assessing the authority to the site and to understand your site, so it will no longer be the case of creating a page stuffed with keywords (not that I ever did that) but to look at what your site is offering.
Content should be descriptive but also serve the user then start to look how the authority is flowing through to the site not just external links but internal linking also. But be careful the machines are also learning on how this authority is being obtained.
Basically, Machine Learning is the future, it’s here to make more of Google’s Users satisfied and so it’s best to really understand it.
As an SEO myself I am up for the challenge.