Brett_Tabke, good to see you here. As I noted, I am in fact counting q/a sites as forums, because they are so close as to be largely indistinguishable. That is, it's kind of an abstraction to say one thing is a forum, and another isn't, when both feature an endlessly flowing set of posts which garner discussion and responses in a dynamic fashion, the main goal being to answer a question.
This is why I consider stack*/reddit forums in this sense, and include their user bases and numbers. We may not like those, I don't, due to reasons actually not that related to their content, but more about how they are run, and how profits are generated by them. If we look at say, the WebmasterWorld HTML/CSS forums, which were basically q/a, plus some discussion if I remember right, but mainly q/a, you'd have to exclude all user posts that are questions by this definition, which is why I include stack*/reddit as forums, since it's too arbitrary to exclude them.
One reason i'm interested in this is specifically to catch up, so thanks for the reminder on panda and penguin, however, though I will not pretend that I am following google seo stuff the way I used to, when it was a bigger part of my work, I still have enough background to be an educated observer of what they are doing internally now, and, again, I have not followed this so I'm probably beating a dead horse that has hopefully been talked to death, since the more mainstream tech media has been talking about the google issue quite a bit, and to me, it's quite apparent that a fundamental core change has happened with google, similar to Tesla FSD beta 11.x and coming 12.x vs their current self driving stack running as 'autopilot'.
What I see is not a small tweak, it's a massive screwup of their core engine, which suggests to me that it has been replaced by something much worse, but on average they hope more profitable to them.
As I indicated above, you simply cannot discuss the decline of forums without directly addressing xrumer and it's ilk, I see this over and over on all the small specialist forums I use and frequent, posts get eaten up by overly aggressive spam filters, required to keep the moderation load down, but destroying, literally, the core functionality of these forums. The problem is, the xrumer devs are aggressive and good, and you cannot use an open source solution to detect them because they will simply add an update to their logic the same way recaptcha was cracked over and over within often hours of the next version.
But because I always prefer testing claims to stating x or y is the case with no data, I will now run a few searches where I know the answer is on forums, some I will not show the key words used but you can trust me that I am not faking it.
Searches are being run on a non logged in chromium, that's the free software version of google chrome, with the google call home bits cut out. I have not done this test before, and these results are the first I have seen of them. In other words, I am not cherry picking, and I am not excluding searches that didn't show what I wanted.
search: keyword1 keyword2: forum result number 4, and the top 3 are all correct. keyword1 is a primary item, and the forums are its forums, but it is not a forum, and keyword2 is something somewhat connected to that but not part of it. This serp result is correct. It's what it should be.
search: keyword3 keyword4: again, keyword3 is a project, and keyword4 is something related to it, but not it. Result one is mint forums. result 2 github, result 3 is the keyword3 website, result 4 is a linux wiki, result 5 is askubuntu, an ubuntu q/a forum, result 6 is reddit, result 7 is debian user forums, result 8 is the creator of keyword4 corporation, results 9,10,11 are technical resources, linux, etc, result 12 and 13 are my forums, result 14 is phoronix, result 15 is another linux forums, result 16 is stack overflow, and so on, I see linux.org forums, antix forums, folding forums, hardforum. Since I know this topic area, these results are correct and quite accurate. I would not argue with anything more than relative positions of wikis vs forum types, and those all seem fine to me.
So quite clearly, for non money keyword searches, google is not blocking forums at all, and somewhat to my surprise, is actually listing them above stuff like reddit or stack* in many cases.
Search: keyword5 tricks: this one actually really surprises me, it's a bit of software and one word that is totally generic. I did not expect my ancient forums and that ancient post to pop up, but it's number 9. The others are how to pages, all correct. So really, you have to dump the notion that google is excluding forum results, it's not.
Now with ones not connected to me:
search: zebra light review [that's a high end LED headlamp, and can be considered a light money key word]: results 1-3 are reviews, as I requested. Result 4-5 are candlepower forums, which is the leading forum in this area. I want to note further that if you get into higher end LED lights, you will end up on candlepower forums, because they are it. That's how I found them.
These results are correct, and google is not excluding any forum resource, particularly since I asked for a review, which skews the results towards single page review pages, not forums. But candlepower also does reviews, so that makes sense. These are correct results, and do not exclude the forums.
Search: foobar2000: results: 1: foobar2000 website, 2, 3, google and apple app stores, 4: foobar2000 reddit, which is basically a foobar2000 forum, 5: wikipedia, 6 majorgeeks download page, 7 hydrogenaudio foobar2000 forums, they actually have a few results in the top 10, as they should, since that's where the developer hangs out. These results are correct, and do not discriminate against forums at all. Note that's a single keyword, with no qualifiers.
This test is for Manjaro linux forums: [
forum.manjaro.org...] and we'll be looking at a few recent topics
Search: manjaro no sound: result 1-30, manjaro forums. Only when you get to 31-32 do you hit a stack* result, then another forums, then reddit, then superuser.com, then framework forums. Basically all the results on the scrolling page 1 are forums.
Search: slackware 15: nothing else, that is, their recent release. Top results slackware site, then some tech review sites, then linux.org forums, reddit, some more tech sites, then linuxquestions.org forums. So no, there is no discrimination in these searches against forums, quite clearly.
So I think we can start to toss out google algo for specific topic areas as reason for forum traffic dropping, clearly.
And again, I do not consider that is valid to distinquish between WebmasterWorld type format of forums and reddit/stack* style, having one be forums, and the other not. That's to me very arbitrary and really kind of subjective, and certainly doesn't fit with reality. For example, manjaro forums are q/a by design, they are there to help users. I hate their forums, as an aside, and I hate that entire js built form thread constructor, but they are just modern spins on forums, which I think yougner people feel more comfortable with.
I remember I used to do these tests against claims a lot as a way to learn, so figured, why not do it again just to see if hte claims being made are correct. Clearly the causal attribution is not correct, though equally clearly, the drop in traffic on what we could maybe call web 1.0 style forums is correct. But drop in traffic to post your issue and get response in a threaded or stacked format from other users of the site is not suffering whatsoever from google discrimination, again, note that in many of the above tests, the forums ranked ABOVE the reddit/stack* results.
So that's not the cause, obviously.
There's some basics that really can't be ignored, as we've learned over years, you have to make the site look current, which is the entire point of making html / css sites. WebmasterWorld does not look current, and needs polishing, because first impressions only work the first time, which is how you get new users.
I liked seeing the raw counts, but for say, WebmasterWorld, that ignores the simple fact that most web searches are hitting stack* sites now, and that's where people are trying to get to, that's what I am looking for often, and when I hit a really unique issue that is so bizarre that I figure someone else should be helped by my posting it, I post it there, and they get upvoted steadily over years. I pick that property because that's where I find many of my solutions (particularly for database stuff, but also some more obscure linux stuff) for work and other projects.
If I wanted to wake up my forums and make them less dormant, I'd start posting new topics there, and those topics would start to rank, same as any time, I just don't do it because I got bored and now they are only active for one main project, which isn't mine. But that project is active, and gets new posts and issues all the time, very technical of course, but users actually seem to prefer as a rule posting on those forums to posting an issue on github for that project.
But there are basics, if the look feel isn't reasonably modern, which is just a css refresh, then users will tend to leave. I tried to warn the perlmonks guys about this issue a few times but gave up, again, it's just a basic template/css refresh. I did that on my forums over a few years over 10 years ago, and they still look fine to me, responsive, all that stuff, it took a bit, but I only had to do it once for the main refresh.
As we all hopefully know, you probably have under 1 second to form a first impression, and if you are looking for new users, you'd better be sure that first impression is what you want it to be, if it's not, there's nothing you can do after that point, unless you have absolutely killer content that nobody has.
These above tests confirm what I have seen now for about 20 years, focused area forums are presented to me by google, they are not hidden or taken out of the serps, I find them, always, consistently, and the winners are the winners, and that's about it. But excluding huge numbers from forums like reddit/stack* is really just cherry picking, there's almost no difference literally between the format of manjaro/garuda/endeavor forums and stack* forums. Yes, forums in the traditional sense are more open and foster more conversation and dialogue, but these are just different forms of the same core concept.
As I indicated, android phones run the linux kernel, and are thus technically linux devices, making linux easily the most dominant global platform in the world, and linux servers run most of the world's websites, and all of their advanced supercomputer and heavy lifting datacenters, basically all of system on chip systems like raspberry pi, but never really breaks 2% desktop market share, unless you include the locked in google chromebooks, which I believe now are built from Debian, used to be Gentoo if I remember right. In the same way, a question and answer site is simply a subset of the larger forum class, it's not not a forum, it's just a big chunk of what forums did, the chunk people seem to lean towards nowadays.
So I would suggest that first, the conclusion that google is excluding forums is demonstrably false, since I just demonstrated it to be false, and second, excluding subsets of forums is like excluding android or chromebook or servers from linux userbase because they don't run the full free software stack, particularly not the desktop and graphics server.
With all this said, I, like most of you, prefer the traditional forum format, but I'm old, and grew up with it, and like it. That's all. I don't pretend however that other forms of it don't exist and aren't also forms of forums however, because I consider them to be just that, someone, or me, posts something, then it gets answered, ideally with a fairly complete discussion about the question. Sometimes these happen on traditional style forums, sometimes on the more restricted q/a forums, but it's the same thing as far as regular users with no skin in the game are concerned, which is why I'd suggest excluding the biggest user group a priori is simply trying to prove that something isn't the case is the case by redefining the thing that is being measured.
As I noted above, I view the biggest issue on the forums I frequent and have seen for years to be the aging out of the users, and failure to replace them with younger users, but that's web style and fad, nothing more, it's like pining for the old days of web portals or something to me. But then I come to new forums, that I am not familiar with, and am blown away by the skill levels of the people there, really top rate, as good as industry, or better, often. This is certainly the case for music type forums, I don't mean fans, I mean the tech, hardware, recording, etc, amazingly good stuff, so I just don't see any issue at all to be honest.
I'm not actually trying to argue this, but I think by defining out the reality and then saying it's not there, that's not really showing anything at all, it's just defining the bulk of users out.
Speaking just for WebmasterWorld, most of its users have been taken by stack*, it's not complicated, that's where they are, they aren';t gone, they are still there in the same numbers, probably much more, they just aren't on WebmasterWorld.
In the old days, I would have decided to start testing this on my forums, but I'm honestly just lazy now, but I think I largely know how the tests would run, I start posting routinely, with quality technical key word posts, google finds theem, and starts serving them up, and my numbers start going up again, never been rocket science, never will be I suspects. Same goes for a blog I have run for ages, if I want it to start ranking again, I just have to start posting content again, in fact, I just did that to help someone I really like get their business up and running re google, and it worked right away, they are top for their local specialized search term, took only a few weeks for that to happen.
But there are also realities, say, for html/css, how many times do you need to learn how to do html css? not many, 1x is enough for most people, then it's just refreshing a bit every 5 years or so. Where or how you do that process isn't very important when it comes to learning it, or refreshing something, or catching up to current stuff once it's stable, you don't need to maintain a continuous forum presence to do that, same for seo etc, once you roughly get it, you don't need to talk about it all the time, just when it becomes relevant, different from learning it and trying to get a working understanding when you get paid to do it.