393 replies to this topic

[Steven UK]

Hi Everyone,

I am quite frustrated at the moment, because unlike any software I have used before, the rankings for any threads being created are just not static.

I'll give you an example. If you have a WP Blog, created a thread, then it gets indexed, and search engines will position the page in the SERPS. Ok great, you think, are although you know the Google dance will affect things now and again, you pretty much know that the thread will be almost static.

Now, with the IPBoard software, forget the Google dance, it has almost become the IPB dance.

For example, a thread will be created, it will get indexed, and stay there for a few days, and if there is a tag associated, then a few days later the tag will leapfrog the actual thread, a week later the thread will vanish to page 3, and the tag will stay on page one, then a few weeks later when Google spiders have crawled the site again, you get pages like the following being indexed INSTEAD of the page that is going to drive visitors to your forum:

http://www.makemoney...&search_app=ccs

The above is an actual page being indexed. Why would page above be getting indexed, and also I have found 'search pages' being indexed too, above normal threads, and the search pages have zero content, and should not be getting indexed at all, really they should not.

One of the other issues, is Google are STILL indexing strings of code onto the end of threads, and because these are not front end threads, Google push them down to page 3 - duplicate content I think the problem is here, which the last update was supposed to fix.

Here is an example of the string: /page__st__20 Now, how can Google even see this? But it does, and it indexes it ABOVE the actual thread, but definitely demotes the importance of the thread due to this issue.

It is a gripe I suppose, and the gripe is there, because this IS fantastic software for ranking, but it is not static. There seems to be so many loose ends that search engines do not know what to do with it, and it seems worse since the last update. Ideally, the ONLY pages that should be getting indexed, are the categories, threads, tags, and that is it. All these other loose ends (quite a few of them) need closing on the IPB software, and this would then be a frightening piece of kit for real marketers, but for now, it is literally a lottery on what gets indexed and ranks, because the software is giving search engines just too many reference points to gather the information, usually the wrong reference points.

If Matt could also take a look, as this is really an SEO issue, that would be great.

[Matt]

I have considered pointing the canonical links on pages to the root page but I don't think that would be right and you'd potentially lose serps on paged items which may contain valuable keywords.

What I want to do is return to query strings for pagination and variables. So you'd have

forum/topic/123-check-this-out/?st=20

I think Google will be happier with that as it will quickly see that even though it's a unique page, it's an 'option' of the root page.

Naturally we'd need to 301 the current /page__ methods but I don't think that is a massive barrier.

[Weppa333]

Good lord this is the RIGHT decision Matt
Can we have this like... now ?
Like a patch or smth ? :!)

[Steven UK]

Matt, what about these type of entries being indexed as a preference by the SERPS:

http://www.makemoney...&search_app=ccs

The above just should not be in the equation really, and neither should search results. How can Google spiders do a search? Not moaning, mate, well... haha.. But the way the software is set up IS most definitely damaging SERPS results, no question. I have put considerable testing into this now, and there are so many loose ends, which I am hoping you guys will fix, but you will only fix them if YOU think there is a problem.

I hope so, I really do.. because all the hard work we put in is being undone on a daily basis.

Thanks for reading the thread.

Edit: Here is a perfect example of what I am talking about at the time of writing (SERPS may change this depending on when you view this thread)

http://www.google.co...lient=firefox-a

Check the 2nd and 3rd entry from our forum, the second one is tags, and the third one is a search entry (stating no results found on a blank page), with the original thread nowhere to be seen.

[stoo2000]

Matt, what about these type of entries being indexed as a preference by the SERPS:

http://www.makemoney...&search_app=ccs

The above just should not be in the equation really, and neither should search results. How can Google spiders do a search? Not moaning, mate, well... haha.. But the way the software is set up IS most definitely damaging SERPS results, no question. I have put considerable testing into this now, and there are so many loose ends, which I am hoping you guys will fix, but you will only fix them if YOU think there is a problem.

I hope so, I really do.. because all the hard work we put in is being undone on a daily basis.

Thanks for reading the thread.

Edit: Here is a perfect example of what I am talking about at the time of writing (SERPS may change this depending on when you view this thread)

http://www.google.co...lient=firefox-a

Check the 2nd and 3rd entry from our forum, the second one is tags, and the third one is a search entry (stating no results found on a blank page), with the original thread nowhere to be seen.

The search box and other elements are removed for search bots in 3.3, you need to update your customised skin to introduce the new features added in 3.3

[realmaverickuk]

Good afternoon guys, first of all, you need to consider the implications of allowing tags to be indexed. Lets say a user creates a thread called "Blue and green widgets" and tags it with "Blue and green widgets". You now have a thread and a tag that are competing for that keyword, which in itself isn't ideal. Though luckily, the tag pages are not optimised, i.e no proper title or h1 tag, which will reduce the likelyhood of tag pages competing with threads.

If you've used Wordpress and either of the main SEO plugins, you'll notice both come with an option to noindex the tag pages, this is generally a good idea. Unless you're really careful and smart using them.

Google are serious about duplicate content these days, though their penalising for it, is quite ridiculous. Historically Matt Cutts told us to simply let Google decide which content to rank, but now post Panda, he's advising us to either noindex, or create a canonical tag.

In this case, the canonical tag isn't the right way to go about it. Canonical tags should be used for stuff like sorting options etc. In this case, you really need to noindex your tag pages, by doing so you'll not get traffic from them, but instead you'll not have tag pages competing, you'll be removing duplicate content and you'll likely see a rise in overall traffic.

Talking about SEO, you really need to do something about this chunk of text.

Please Read: How The Forum Works:

The Make Money Forum is a forum dedicated to working from home, making money from home, and much more. This area is a large topic of debate on the internet, and we aim to provide unbiased information in a way that can benefit the majority of the members who visit. Any reviews given by forum administrators or moderators are not a definite indication of fact, and will be provided in a way to help viewing members to make informed decisions during their research into the mentioned subject. More and more people are looking to work from home, in an online business and make money forum gives advice on a variety of different solutions, and support when looking to create home incomes. Join our forum today, and start making money online today, in an array of different opportunities to make money online.

If you have a genuine interest in these areas, then we provide a platform to discuss companies you are interested in, and maybe who you have used, with the purpose of helping other members. The forum is not a venue for members to advertise their wares. If members need to contact the staff of the forum, then the link to do this is provided at the top of the page. Please note, that the forum is not responsible for member's own views, and does not take any responsibility for postings from members.

That's quite a chunk of text and it's appearing on every page. That's a serious amount of duplicate content right there. An easy fix, is to turn that chunk of text in to an image. A single duplicate image, won't be nearly as bad as what you've got there.

The text is also quite targeted with keywords such as "Make money online", which will appear like keyword stuffing.

[realmaverickuk]

I have considered pointing the canonical links on pages to the root page but I don't think that would be right and you'd potentially lose serps on paged items which may contain valuable keywords.

What I want to do is return to query strings for pagination and variables. So you'd have

forum/topic/123-check-this-out/?st=20

I think Google will be happier with that as it will quickly see that even though it's a unique page, it's an 'option' of the root page.

Naturally we'd need to 301 the current /page__ methods but I don't think that is a massive barrier.

Matt, this isn't the right way to go about it, it would definitely be misuse of the canonical tag. Remember that the canonical tag is to set a preferred version. Page two isn't a version of page one.

You need to be super careful how you use this tag, it's potentially lethal.

What you could do, is look at the major tags on the page, namely the <title> and <h1> the title tag is appended with "page 2", but more could be done to switch it up, to help prevent page 2 from competing with page 1. For example:

Page 1: Title: SEO Rankings flying all over the place | Invision Power
Page 2: Title: Page 2 of 2 | Invision Power | SEO Rankings flying all over the place

This would also push the target keywords, closer to the end of the title, meaning it gets less weight, helping page 1 become the clear authority for this keyword.

Something similar, could be done with the page <h1>, though slightly different methods, as we need to take usability in to account also.

[realmaverickuk]

Here is an example of the string: /page__st__20 Now, how can Google even see this? But it does, and it indexes it ABOVE the actual thread, but definitely demotes the importance of the thread due to this issue.

Steven, you need to also understand that /page_st_20 is simply page 2, which SHOULD be indexed. You will get traffic, for actual posts made. Lets say on page 2 I make a post that within it says "I've been making a ton of money online....". The thread title is "How to make money". And somebody searches Google for "Making a ton of money online", the thread has the potential to rank, based purely on the content of my post, on page 2.

The issue isn't that the posts are indexed, it's that they are too similar to page 1 and compete with it, as you're seeing. The above suggested fix, would likely do the job and prevent it from ever competing.

Of course, /page-2/ would be a much more universal language for Google to interpret vs page_st_20

[Steven UK]

The search box and other elements are removed for search bots in 3.3, you need to update your customised skin to introduce the new features added in 3.3

Thanks for this info. Please could you explain how we would go about doing this?

[realmaverickuk]

The above is an actual page being indexed. Why would page above be getting indexed, and also I have found 'search pages' being indexed too, above normal threads, and the search pages have zero content, and should not be getting indexed at all, really they should not.

Agreed, search pages should be noindexed IMO.

[Steven UK]

realmaverickUK, I would honestly prefer those strings not to be there, and the pages to remain static. I have run forums for over 10 years, and never come across that situation anywhere else with any other software, and it is detrimental to the searches, it confuses the hell out of Google, because it is just not static. One day the original thread will be there, 2 days later Google crawl again, and the original thread moves from page one, to be replaced with a {string} added to the URL, that has now been pushed down to page 2/3. 2 days later, it gets replaced by something else.

My other websites I own (120+of them) do not do this. If I create a blog post on WP, it stays where it is, because WP does not offer the SERPS any alternative, other than tags, which you have to be clever about, how you use them.

You are right though about the tags, they definitely do compete with the main thread, but with IPB, not only do the tags compete, you have other variables too, competing against the main thread. I have actually, as of yesterday stopped using the tags, as they are destroying thread rankings, which I take some responsibility for, for asking Matt if they could be indexed. I just did not realise that they would compete so much with the main threads, as I have never experienced that with any other software.

The search box and other elements are removed for search bots in 3.3, you need to update your customised skin to introduce the new features added in 3.3

Can somebody please advise how we do that??

Thank you.

[realmaverickuk]

Thanks for this info. Please could you explain how we would go about doing this?

Steven goto your template and open globalTemplate

search for

					    {parse template="quickSearch" group="global" params=""}

Replace with

	    <if test="canSearch:|:$this->memberData['g_use_search']">
					    {parse template="quickSearch" group="global" params=""}
	    </if>

This basically hides the search box from Google and other engines.

[Steven UK]

Thank you.

[realmaverickuk]

realmaverickUK, I would honestly prefer those strings not to be there, and the pages to remain static

Ideally, it would say /page-2/ which all of your websites will do. So to remove the query string, rewrite the url's for paging.

But the issue is NOT the query string. The issue is that page 2, is competing with page 1 due to the almost exact match <title> and <h1>. I outlined the fix for that above.

VB, Wordpress and every other CMS on the planet, all have paging. Paging is fine.

Just be very careful what you ask for. Because had Matt gone ahead and added canonical to page 2+, you'd have ended up in a bigger mess.

I don't really use the forum side of IPB, and I've never fully explored how it can be optimised, but it's clear there is a lot to be improved but it's good that Matt's listening.

All of the advice I have posted here, isn't particularly well thought out and with any SEO plan, the implications need to be considered properly and the plan needs to be solid.

At present, Google is confused, it wouldn't take much to fix it all, but it would need to be well thought out.

[realmaverickuk]

Something else, food for thought.

When you use tags, the thread itself, links to the tag page with the anchor text "keyword", you're telling Google, hey go visit this page, it's about "Keyword".

Then, if you've got a page 2, 3, 4, 5, all of these pages also have the anchor text "keyword" linking to that page.

Then if somebody else creates a thread, and also adds a tag "keyword", thats a whole load more links pointing to the tag page for "keyword".

TBH, it's not hard to see, why your tag pages could outrank your main page.

[Steven UK]

Oh of course, page 1, page, 2, no problem with that, it is the way it is being read by the SERPS currently that I believe to be the issue.

Agreed, the pages are competing, when they should not be, they should just be an extension of page 1, and be read that way, and then usually what happens is that the SERPS will place page 2, under page 1 right underneath each other.

So currently, when you throw the tags into this mix, then suddenly you have the tag, the main thread page one, and all the strings (which need rewriting to pages 1,2, etc.) competing, so the answer is:

So the fix would be to rewrite the url for paging.

To be done at IPB level in a patch of some kind, that would be great.

TBH, it's not easy to see, why your tag pages could outrank your main page.

I have never come across that happening with other software, but it is becoming the standard now on IPB (probably not by design, but still, it is fact) , that as soon as the SERPS find the tags on IPB, then bang, it throws the main thread to a lower page, and keeps the tag on the first page as a preference. Strange.

[realmaverickuk]

Oh of course, page 1, page, 2, no problem with that, it is the way it is being read by the SERPS currently that I believe to be the issue.

Agreed, the pages are competing, when they should not be, they should just be an extension of page 1, and be read that way, and then usually what happens is that the SERPS will place page 2, under page 1 right underneath each other.

So currently, when you throw the tags into this mix, then suddenly you have the tag, the main thread page one, and all the strings (which need rewriting to pages 1,2, etc.) competing, so the answer is:

To be done at IPB level in a patch of some kind, that would be great.

I have never come across that happening with other software, but it is becoming the standard now on IPB (probably not by design, but still, it is fact) , that as soon as the SERPS find the tags on IPB, then bang, it throws the main thread to a lower page, and keeps the tag on the first page as a preference. Strange.

I agree it's an issue, but the problem nor fix, are related to the /page_st_20, that's just another page, like /page-2/

All that we need, it to effectively un-optimise page 2+, for the target keyword. As outlined above, is a really simple solution.

Page 1: Title: SEO Rankings flying all over the place | Invision Power
Page 2: Title: Page 2 of 2 | Invision Power | SEO Rankings flying all over the place

This would prevent page 2, from competing with page 1.

The reason the tags are competing is explain above, regarding the anchor text.

It's all fixable, but as I said, it's important to plan properly and consider ALL implications of changes like these.

[realmaverickuk]

You quoted my post, before I changed the "It's not easy to see", to "It's not hard to see".

Hopefully Matt can address the issues, but before he does, I'd like some time to rethink everything I've suggested to double check it's definitely the way forward. But you may have to resort to hiring an SEO and IPB developer, and create fixes and optimisations yourself. But if it's a profitable website, it would be worthwhile.

[Steven UK]

Google would usually rank the pages by date, so page 2 should automatically go UNDER page 1, with a date attached, that is usually how they filter.

[Steven UK]

You quoted my post, before I changed the "It's not easy to see", to "It's not hard to see".

Hopefully Matt can address the issues, but before he does, I'd like some time to rethink everything I've suggested to double check it's definitely the way forward. But you may have to resort to hiring an SEO and IPB developer, and create fixes and optimisations yourself. But if it's a profitable website, it would be worthwhile.

Maybe, but I believe IPB HAVE to implement these changes, it is basic forum characteristics we are talking about here, that help SEO. Tags? Yes, they always compete, but NEVER to the extent that IPB tags do.

Seriously, I have never in all my years of ranking sites seen so much fluctuation amongst results, than IPB.

This is how the rankings should look for pages related, page 2 or not:



They place them in date order, even if it is the same thread, they just use page2, page3, page4, then date it. This does not happen with IPB. There is simply no relationship between anything, the threads just compete, and by the way, this page 2 being mentioned, the strings are still being product even if there is a single page, most of the pages on my forum are single pages, that have not overlapped onto another page, yet the SERPS are still indexing a string.