Effects Of Removing 70% of Website Pages?

[EmptyRoom]

Hi guys,

What do you think would happen if I removed 70% of all pages on one of my websites? Those pages are filled with content of lower quality than the remaining 30%, and they don't receive any traffic (minor).

The site's making money, although a lot less than before Panda hit. I am afraid I'll lose all rankings if I remove such a large number of pages.

What do you think?

[Planet13]

I believe there is a post in this forum by wheel in which he asked that same question, so you might try and find that forum thread.

I think for his situation, he didn't have any problems. But try to find that thread if you can first.

[ascensions]

I'm doing it right now, and have had very positive results. Not necessarily in traffic, as I'm down 5% of traffic even after Panada's 50% cut, but the CTR shot through the roof.

[g1smd]

I've just cut 50 000 pages away: low content, and duplicate content issues.

A combination of 301 for moved pages, and 410 for pages permanently gone, has worked wonders.

Google has taken many months to take it all in, and I guess it will take another couple. Nearing the end of it, overall traffic and page views per visit are all up.

[ascensions]

Panda will be known as the "Great Purge" by our descendants...

[Karma]

I'm considering either a delete or noindex on mine, although I'm dubious about the benefits at this point.

[EmptyRoom]

I'm doing it right now, and have had very positive results. Not necessarily in traffic, as I'm down 5% of traffic even after Panada's 50% cut, but the CTR shot through the roof.

How do you explain this? Why did the CTR increase?

I've just cut 50 000 pages away: low content, and duplicate content issues.

A combination of 301 for moved pages, and 410 for pages permanently gone, has worked wonders.

Google has taken many months to take it all in, and I guess it will take another couple. Nearing the end of it, overall traffic and page views per visit are all up.

Why 410? Why not 404?

[Karma]

I think 410 is for pages that once existed but have been removed, whereas 404 is for pages that have never existed.

To Google at least, 410 is apparently considered more permanent than a 404.

[seroundtable.com ]

[g1smd]

404 is a generic "the server didn't find what you asked for, doesn't know why it isn't there, doesn't know if it ever existed, and doesn't know if something will be available here at some later date", whereas 410 is the more meaty "this stuff has gone and it ain't coming back; don't bother checking again".

Google supposedly treats them both a little differently.

[Karma]

What about Bing/Yahoo?

[ascensions]

@Empty Room... I can't explain it... I purposely cut pages that didn't make money, so they would have factored very little into the CTR... It's not just the CTR that's gone up but the value of the clicks too. It's almost as if Adsense knows the site is smaller, despite it not showing a size difference in the index. Eventually I suspect PR will flow different, but it was almost an instantaneous effect. I think it's best to begin saying, Google now works in mysterious ways.

[Frost_Angel]

I took a hatchet to my site after I lost 80% of my traffic April 10th. I removed an article directory for fear of duplicate content, a business directory because of low traffic, my entire forum just out of fear alone and about 30 pages of content. I also noindex-ed about 20 pages because they were free printables and just images.
Has is done a damn thing for my rankings? NO.
My site ranked well for 11 years in it's niche - I wish I hadn't freaked after I got Pancer. Just my 2 cents.

[wheel]

I removed about 5-10K pages from my site, pre-panda. THey were weakly linked to but had exceptional content.

No measurable difference to me. Certainly no big drop in the rankings.

[optimierung]

@Karma: 410 has the effect that the pages are disappearing much more quicker from G-index

[PPC_Chris]

After being hit by Panda on April 11, I removed 90%+ of pages that could be considered thin content and have not seen one bit of recovery.

[indyank]

If this is done to recover a panda affected site, then it might not help. It does require a manual intervention from their end to release a panda flagged site.

[MrSavage]

My thought is this. If you remove those pages in hope of Panda recovery, consider your feelings if Bing was suddenly the #1 search engine. Ask yourself this, are you doing it because you think Google wants that? If the answer is yes, then think very hard before doing it.

In my opinion, and nobody has insider information about Panda, I would not delete anything. That's your work, your time after all. I would refile it via 301's. Either a new page with say 10 pages of thin content onto that new 1 page, or find a similar type page that exists and paste the content on there. Use the 301 to tell Google that you moved the content.

I can only say that my experience of chasing the pot of gold (Google penalty reconsideration's, Google algo changes, Google Panda) that at the end of the day it's a black hole. It's a horrible feeling. With this Panda situation it's worth merging content, and if it's not complete duplicate, I would be cutting what I can and putting onto existing pages and doing a 301. That's it. Makes my site better imo and there is an off chance that Google might like it enough to lift Panda from your A.

[freejung]

It does require a manual intervention from their end to release a panda flagged site.

That's huge, and the first I've heard of it -- where does that info come from, if you don't mind my asking?

In that case, the question becomes one of how to get Google's attention enough to merit reconsideration.

That's terrible news for smaller pandalized sites, it could take quite some time to get around to reviewing all of us.

[bwnbwn]

1st I heard as well. Panda is not a penality but a ranking change due to factors, I can't see it being a review, unless i missed something.

[indyank]

Google wouldn't be making that info public. But it is not that difficult to get to it, if we could go through all the statements made by each of the googlers, since Panda was released.

There were a couple of statements by two Googlers. But the one made by their webmaster trends analyst several months ago does almost suggest that. He did say the usual things like "improve quality", "focus on users" and so on but he did add that the recovery would take its own time and it might not happen that quickly and automatically.

After all this is a manually run process to flag the low quality ones.

I was reading those statements from my notes of the early days of panda and this particular googler's statement was revealing.

I will have to add that I did see a recovery to one low traffic entertainment site but it might not have been a panda victim as the imp. dates doesn't confirm it.

[tedster]

Here's the way I understand Panda. Changes in any site's Panda score currently happen when Panda is re-run over all the sites. When to execute that re-run is a manual decision - that is, it doesn't just happen on an automated schedule.

However, when Panda actually is re-run, then all the new evaluations are automated for every website. In other words, individual websites are not manually re-evaluated. The manual part is Google deciding to re-run Panda.

[AlexB77]

We have already heard that some sites managed to recover and there were nothing mentioned about manual removal from "panda box". I simply would not suggest making such statements here if source is not disclosed.

[AlexB77]

I will totally agree to tedster

[freejung]

Ah, OK. I have to say, I'm extremely skeptical. It probably takes manual intervention to start a Panda update, and that may be what the googler in question was referring to, but I seriously doubt it requires direct manual intervention to change the "Panda score" of a particular site. That would not be consistent with Google's MO, nor would it scale, nor would it even make sense. Why bother automating the process at all if you're just going to have to manually review every site it flags? I don't buy it.

Other than that, I agree with your original statement - if you're removing content to recover from Panda it may not help. Not because manual intervention is involved, but simply because that may not be the reason, or at least the only reason, you were pandalized in the first place.

[indyank]

Why bother automating the process at all if you're just going to have to manually review every site it flags? I don't buy it.

But not many sites recover after each of those manually triggered runs. This whole process is still manual not just because of the intensity of the algo but it is still in the early stages and a lot of learning is involved.

If a bunch of sites come out positive at the end of a run, they would definitely want to make sure they have sufficiently changed to become positive.

[freejung]

If a bunch of sites come out positive at the end of a run, they would definitely want to make sure they have sufficiently changed to become positive.

I'm sure they would take a random sample for QC purposes, but it seems unlikely they would review all of them.

[potentialgeek]

Manual reviews were used to set up the algo, but they aren't used to get a site back from Pandalization. Given the very small number of reported recoveries, however, you can speculate how Google could have done manual reviews.

This is a great thread because I have a site with about 2,000 pages that I was about to start changing. I am torn between:

a) deleting;

b) hiding from Google;

c) merging;

d) developing with new content;

e) editing;

f) complete rewriting;

g) starting a brand new website (for some of the content);

h) starting a brand new website (with all/most of the content); and,

i) making the site brand-like (and therefore immune to Panda);

I have not seen enough recovery stories to know which of these nine basic options would likely be the most successful. Google itself has suggested moving weak content or adding new stuff, etc., generally making it unique and higher quality (fit for a magazine).

[MrSavage]

If you can wait a week or so I may have an update for one of my sites regarding recovery. I'm wise enough not to say anything unless this sticks for a couple weeks.

In terms of your list, again just my opinion, merging/301 is what I would look at doing. The reason is, those images and content are your time and effort. It's also the fastest least effort way of moving forward. If you can bulk up existing pages via merging, or create a new page with multiple pages merging that share a similar subject, I say consider that first.

Again, I'm experiencing pre Panda traffic, but I seen this movie before. As in 5 days of glory only to be kicked in the B and dropped lower than ever in Google results. So I will report back on my situation which might help your decision making.

[tedster]

It might go without saying for many people, but I would emphasize that if you have "content" where you know you cut corners in some way, that should be addressed - either fill it out to a good quality level or go 410 in bad cases (scraped/spun/syndicated stuff).

[freejung]

MrSavage, that would be quite useful. I'm in a similar position to potentialgeek, and I'm currently mostly doing d) and e). I've considered doing some radical merging - i.e. combining entire categories of pages into a single page - but I'm hesitant to do that for several reasons.

It's also the fastest least effort way of moving forward.

That's what worries me about both merging and deleting. It's too quick. The more I look at it, the more I think Google has deliberately designed the Panda algo to be resistant to any sort of quick fix. That would make a lot of sense to me, as the problem with "content farms" in the first place is the quick creation of low-quality content. I'm guessing that unless what you're experiencing is just a technical issue (dup content because of CMS errors for example), anything that doesn't take many hours of work is likely to be insufficient.