Here the cutting and bleeding edges of marketing, analytics, conversion, accessibility, usability, findability, user experience et al can be discussed without fanaticism, building from all the past members' commentaries within the archives. I find it fascinating to follow the threads of thinking from those beginnings to now; so little has actually changed! Yes, details are different, things are certainly more complex, but the various principles pronounced here at Cre8 have remained surprisingly robust and current.
Yep. You're spot on, iamlost. Here's a nice Guided Tour of some of the old SEO discussions from the olden days - and their relationships to what is going on right now. (I chose to tour that section because it's the section I most participated in, and therefore, I remember most of these threads from when they first came out).
Let's start with Ammon's Quick Kick Start Guide to SEO., For years, that was stickied and probably the most important piece of reading for any new SEO interested visitor to the site. It's no longer stickied, because most of the links to various resources are long gone, but the advice remains true.
Last year, Google introduced RankBrain to their Hummingbird Algorithm. This has had people confused and freaking out ever since. Only the most hardcore of SEO people really seem to understand what is happening in all that - and even then, many of them are afraid to profess any certainty on the subject. But, if you've been playing along with us at home, you've known exactly what is going on since 2002 when Ammon started a thread which got no real attention for one year and five days. That thread was aptly entitled "The Semantic Web." This concept differs slightly from RankBrain in that it is about looking at things from the perspective of the search terms, while our observations were about the things they were doing at the time to focus from the perspective of the search results.
In February of 2003 we started talking about LocalRank. It was at this point where we had (almost) all of the ingredients needed to understand what RankBrain is trying to do. It takes linked clusters of things and tries to establish relationship patterns - an integral part of the Semantic Web concepts. It was here that I started to tell folks that outbound links could be as important as inbound links because of their power to establish relationships with similar and/or related concepts.
In November of 2003, all of this stuff started making its way into Google's code. Barry Welford was the first to spot that Google is Stemming. Both Ammon and Peter D alluded to the fact that there was a bit more going on that any real "stemming" but at that point - everyone was being a bit quiet until they knew for sure what all was in play.
Two months later, we knew more and started to try to explain it when Kim asked us the Difference between Stemming and Semantics.
Now, 13 years after all of that, people are freaking out about RankBrain and how much of a HUGE change it is and how it changes everything. The real truth is, the only thing truly NEW about RankBrain is the fact that it is being allowed to learn by itself without user intervention (that, and its shiny new name). It's all been happening for over a dozen years - it is just doing it well enough to be trusted to learn on its own.
In 2016, Danny Sullivan wrote a RankBrain FAQ in which he described it as follows:
As Google told us, it can see patterns between seemingly unconnected complex searches to understand how they’re actually similar to each other. This learning, in turn, allows it to better understand future complex searches and whether they’re related to particular topics. Most important, from what Google told us, it can then associate these groups of searches with results that it thinks searchers will like the most.
Google didn’t provide examples of groups of searches or give details on how RankBrain guesses at what are the best pages. But the latter is probably because if it can translate an ambiguous search into something more specific, it can then bring back better answers.
Way back in 2004, I described the "Stemming (but it's not really stemming, it's Semantics)" phenomenon and it sounds quite similar except for a matter of perspective. (And I gave examples, unlike Google)
Semantics, though, looks at words and phrases to see how they are commonly used in conjunction with each other and will also compare them across multiple pages.
Google has a database of "relationships" between pages that shows which pages link to each other and then culls common phrases from those pages and attempts to come up with a semantical meaning for the words in the context of this set of pages. By looking at the OSHA site and all the pages that links to it, Google can see a common term - let's say "Fire Exit" - and come to the conclusion that "fire exit" is related to "Safety".
On another batch of pages, you have the phrase "fire pit" used quite a bit. Because of the other elements on the pages, Google can say, "fire pit" doesn't have anything to do with safety, but rather, "Cooking".
A little further on I speculated that learning would be a part of it:
This relational database that controls all of this grows over time, I would imagine. And it would also adapt to new phrases and industry specific terms as more and more people adopt them. For example, last month, "Florida" was a place. Of late, though, if you see the words "florida" and "update" near each other, we're not talking about a place, we're talking about Google. Google's not going to know that right away, but as it crawls the web and spots more and more examples of the term "florida update" the more confident it will be in making the "florida update" = "Google Update" assumption.
Aside from the fact that 2016 switched perspective to start looking at the terms rather than the results pages, the concepts are almost identical, aren't they? Ultimately, Rankbrain was a simple, small, evolutionary step in the progress of data searching.
These forums always were and always will be the one resource that taught me more about the industry in which I work than any other source. The information may not be flying around as fast and furious as it did back then, but the information is still invaluable. Last year, this place reached 18 and as such, it was old enough to vote. I heard that it neglected to register for the last election due to the lack of a valid choice at the polls, but it's nice to know that it could have.
If we can hold on for 2 more years, this place will be old and wise enough to drink. Now that's something I'm looking forward to.