Post Panda: Splitting a Site, then Interlinking the New and Old

[ken_b]

Google has suggested splitting up a site that has been affected by the Panda Update. The idea being to offload the low quality or shallow content pages so they don't affect the ranking of the rest of the site.

But if you do that, is it a good idea to interlink the newly split sites?

At first glance it seems like a no-brainer to interlink the sites.

BUT... if the idea of splitting the sites is to offload the pages that are dragging the site down, doesn't it make sense that linking to the offloaded site(s)/pages a bit risky?

If the pages sre harming the original site, why wouldn't linking to the on a new site also be considered a negative signal?

Think linking to a "bad neighborhood"?

I suppose you could redirect the offloaded urls via 301, but couldn't that be even nore problematic that interlinking the sites? With a 301 aren't you telling Google that there is a clear and solid association between the old site and the new "low quality" site"?
.

[Shatner]

Google suggested it? I haven't seen anything where they actually suggest it.

[ken_b]

I haven't seen anything where they actually suggest it.

Here ya go ....

Think you're affected by the recent algorithm change? [google.com]

Removing low quality pages or moving them to a different domain could help your rankings for the higher quality content.

[koan]

I vaguely remember reading a suggestion by someone from Google that moving thin or low quality content to its own subdomain may help the original site.

Edit: ken_b was quicker.

[dickbaker]

I've been considering the same thing. I have an ecommerce site that I believe Google may see as not fitting with the focus of the rest of my site, so I'm mulling over the idea of spinning it off into its own site.

In doing so, I plan on still having a link at the top of the pages on my main site that says "Online Store", just as I do now. The difference would be that the link goes to another site, and I add a "nofollow" tag.

If that's a bad idea, I hope there's posts coming in this thread explaining why.

[Shatner]

Thanks ken_B! I'd totally forgotten about that aspect.

If they're going to recommend it, would be nice if they'd provide some guidelines about the best ways to do it rather than letting webmasters guess and get themselves further penalized.

[johnhh]

Been thinking about this for 4 weeks now.

The pros are it may work
The cons are it may not , and if the pages are removed completely all the happy visitors that have bookmarked pages suddenly find they don't work anymore , so the only option is to 301 the pages to a new site/subdomain, and it's a lot of pages so what will happen then is anyone's guess.

Another option: if all the identified "weak" pages are noindexed and just found by natural navigation of the site that would work for us but, of course, Mr Google already knows about them.

Either option would result in a long wait and probably only be worthwhile for sites that have some merit.

Back to thinking.. as some competitors have been hit and others, selling the same products and similar site structure, have not.

[browsee]

Removing low quality pages or moving them to a different domain could help your rankings for the higher quality content.

Are they really talking about subdomain? It is really confusing, there is no clear direction on subdomains.

[tedster]

I don't think it's confusing. It basically says "get that junk off your site." It doesn't say "try to squeeze all the value you can out of it, even though you know it's crap."

Splitting off to a second site sounds OK to me, if you want or need to keep that content online. But cross-linking sounds counterproductive. Even if it "works" for a while (doubtful, I'd say) it's not likely to work for very long.

Domain names are not expensive, so I wouldn't even think about a subdomain.

[dickbaker]

Domain names are not expensive, so I wouldn't even think about a subdomain.

The reason I was asking about subdomains earlier was that I think doing so would be less confusing to visitors who know the site. If they see ecommerce.mysite.com or information.mysite.com where before they were seeing mysite.com, they likely won't feel lost. If they see ecommerce.com, they may wonder where the heck they are.

[tedster]

That would mean that you intend to link to those moved pages then - right? If you can see that many visitors use and value them, then removal of any kind may not be the right approach for you.

[johnhh]

then removal of any kind may not be the right approach

We have thousands of pages that Panda may see as "weak" - yet are popular pages on the site.

subdomain.example.com may be the way to go , redirecting pages to that sub-domain.
The problem is - will this be enough to raise the perceived quality of the original site given the number of 301's that would be required, or will the number of redirects result in a further downgrading/no effect.

If we attempt a test of this on one of our minor sites it would be many months before a result is known and even then there could be other factors in play specific to the test site.

[apauto]

Why not throw the low quality pages on a subdomain?

[tedster]

If recovering Google traffic is the goal, then I'd follow the public instructions that they've given. But it is your site - so you can certainly do whatever you want.

We have thousands of pages that Panda may see as "weak" - yet are popular pages on the site.

How have you determined this? Are these the pages with the biggest search traffic loss? That's where I'd start, rather than a wholesale revamping of everything you can find.

[Planet13]

We have thousands of pages that Panda may see as "weak" - yet are popular pages on the site.

How difficult would it be to "beef up" those pages?

[dickbaker]

That would mean that you intend to link to those moved pages then - right? If you can see that many visitors use and value them, then removal of any kind may not be the right approach for you.

Yes, Tedster. That's what I'm thinking of doing ("thinking" being the operative word right now).

This goes to what we were talking about in another thread about a site's focus. I strongly suspect that having an ecommerce section on my site is diluting its focus in Google's eyes. Judging from the phrases that still rank relatively well, Google regards my site as an authority on retail stores in certain cities and states, and on retail stores that sell certain brands of widgets. It apparently doesn't regard my site as an authority on the widgets I sell, although pre-Panda I was page one for nearly all of them utilizing I don't know how many different phrases.

I don't know how to test this theory other than just doing it. Right now the customers to that part of the site are coming from Google Shopping, from Adwords, from Yahoo and Bing and other engines, and from the "Online Store" button on the site. They're not coming from Google organic searches.

It's no small decision, so it's hard to pull the trigger on it.

[indyank]

I strongly suspect that having an e-commerce section on my site is diluting its focus in Google's eyes.

Dickbaker, I don't seen any issues in having a forum and blog attached to a e-commerce site on the same domain as long as they are all on related topics. In another thread on forums, a few people have even reported that their forums were untouched by panda, while their main site was pandalized.

If you remember, a few years ago Matt Cutts and Google were supporting the idea of having a blog attached to corporate sites.Yes, I do know that they keep changing what they say and feel about what is good and bad, but I don't see any harm in having a blog or forum attached to an e-commerce site on the same domain, as long as they are on related topics.

What google seem to be suggesting is spinning off the low quality content to separate domains.But this suggestion looks totally illogical.Why do they want people to have separate domains for low quality content? They could have stopped short of suggesting to remove the low quality pages from the google index.

The idea of moving low quality pages to a separate domain was suggested by one of their employees on their forums and I am not even sure why he suggested it, as it doesn't go with what they also advocate - "remove low quality pages from SERPS".Even if domains are cheaper, why should one have a domain for low quality content?

[indyank]

I am also now seeing a lot of people having wordpress sites changing their permalinks and doing mass 301 redirects to escape from panda.These are people who seem to be doing this based on what is discussed here and on google forums.But I am sure that these folks will be hit hard by google soon and 301 redirect is easy for google to detect.

[johnhh]

How have you determined this?

"Widget" descriptions from "manufacturers" .
This is working on the theory that weak pages affect the strong pages by being linked together, even if the link is one way. Hence the pages that lost most traffic may not be the pages with the problem.

How difficult would it be to "beef up" those pages?

Difficult, as in effect pages lists blue widgets, red widgets by different sizes and voltages with a short snippet - like SERPS infact !

Our traffic appears to be normalised around a mean.

[dickbaker]

Indyrank, this is another of the Google theories that's being bandied about, but it's one that makes sense to me. I've suspected for a year or so that the ecomm section of my site was confusing to Google.

To quickly reiterate from another thread: I have a site that has full-page, searchable ads for widget stores, broken out by state, and searchable by dozens of type of merchandise or services. I also have a couple thousand pages of photos and spec's for individual models of widgets from various manufacturers.

Lastly, I have a small ecommerce section selling some widget accessories.

Just about everything I had for all three sections was first page on Google. Panda put my rankings for any of the ecomm phrases down to page three, four five or even ten or eleven. For the widget spec and photo pages, I dropped at best to page two, at worst to page 18, and the rest somewhere in between.

The pages for the ads for retail stores, though, still rank fairly well, as do phrases such as "[insert a state name here] widget stores" or "[insert a city name here] widget stores", or variations of that sort of phrase. The searchable ads were the original focus of my site when I launched it seven years ago, and I strongly suspect that's what Google sees as the site's focus now.

There's more on this discussion in this thread: [webmasterworld.com...]

[tedster]

"Widget" descriptions from "manufacturers"

Google has been clear about how they see this from long before Panda. In fact, it was the origin of the phrase "thin content" several years back.

Google said Panda's target is "shallow content" and they describe that in broader terms than "thin content" that merely reproduces the manufacturer's descriptions or an affiliate feed. Sure thin content is most likely one variety of shallow content.

Manufacturer's descriptions are pretty much a dead duck. You will need to re-think how you offer those products if you hope to get significant search traffic to those pages - and even more, not have them affect the rest of the site's rankings.

[ken_b]

I was hoping for a discussion in this thread about whether it was a good idea to interlink sites that have been created by splitting a bigger site post Panda, not about if a site should be split.

For this discussion I'm making an assumption, not always a good idea, that some split off content has a real value and is fairly popular with visitors, but that it doesn't measure up to the balance of the content and thus drags the whole site down in the serps, so it gets split off.

Given that, if it's not a good idea to interlink the new and old sites, what might be the next best path for getting the new sites indexed? Do you have to go all the way back to basics in link development?

Is there a safe or safer middle ground?

[johnhh]

those products if you hope to get significant search traffic to those pages

We don't expect to get search traffic to those pages - although it is suprising how much we do get through Bing - they are there for the user to go to after browsing unique articles, which we do get Google search traffic to. And of course we would like them to buy a widget !

The point is - are these "weak" pages dragging down the rest of the site by association and is there an advantage in spinning them off into a subdomain.

ken_b: I think you have decide whether it is a good idea to split in the first place - if it is then the problem of linkage occurs.

[NewOrder]

I was hoping for a discussion in this thread about whether it was a good idea to interlink sites that have been created by splitting a bigger site post Panda, not about if a site should be split.

Also whether a 301 redirect should be done from old pages to new ones on the new domain. Surely this should be done?

[ken_b]

Also whether a 301 redirect should be done from old pages to new ones on the new domain. Surely this should be done?

Also whether a 301 redirect should be done from old pages to new ones on the new domain. Surely this should be done?
I seriously question whether the old site and the new site(s) should be interlinked at all.

I ave even more concerns about anything more than minimal interlinking.

Using a 301 redirect on the pages seems like it might be even more risky.

But I also think it might depend on why the site is being split up.

If the offloaded pages are truley junk, well then I suspect they shouldn't even be set up on a new site, just delete them and be done with it.

But if the site is being split into pieces that all have some real value to readers but lead to a lack of focus when on a single site, that's another issue. That's especially true if one section has been classified by Google as being of less value and drags down the rest of the site.

[dickbaker]

Manufacturer's descriptions are pretty much a dead duck. You will need to re-think how you offer those products if you hope to get significant search traffic to those pages - and even more, not have them affect the rest of the site's rankings.

Tedster, what I've been doing with the manufacturer descriptions is rewriting them for freshness, then adding some extras to the page: product reviews by customers, or an "our impressions of the model XYZ" mini-article, or, where applicable, a bit of history on the model XYZ. It takes time to do the latter two, so I hope they pay off.

Back to spinning off my ecommerce section, if you don't mind. I think it's the right thing to do, and not just because I think there's a focus factor with Panda. The problem with spinning it off is that I have very good rankings and ratings in Google Shopping for those ecomm pages. If I go to a new domain, those rankings won't transfer. I'll be starting over again, and right now, Google Shopping is about the only place where the conversions are coming from.

Move the section, and lose rankings in the only place where sales are coming from. Keep the section where it is, and it may be contributing to the demotion of the entire site, including the ecomm section.

Talk about a rock and a hard place.

[ken_b]

dickbaker

Back to spinning off my ecommerce section, if you don't mind.

Might be a good topic for another thread.

This thread is about whether linking to sites that have been split of is a good idea.