Panda's BIG Weakness ! Why sites can't recover

[seoskunk]

I have long suspected this but now I have seen this with my own eyes. Recently I made major changes on a site. New site titles new content and the site collapsed in ranking. Maybe its their new patent causing this, needless to say these changes were necessary for the long term and I'm sweating the short term. However here is the point of the post. As the site plumetted I noticed these changes to overall stats.

1. Bounce rate up from 48% to 75%
2. Visitor time on site down from 3.48 to 0.35
3. Spam increase and Zombie Traffic
4. Returning Visitors down by 30%

The point is the lower the serp ranking the worse the stats are regardless of the quality of the site. And this is where Panda fails I believe. Once your down the serps you naturally get abysmal results and algorithmically is impossible to tell the quality of the site.


Your opinions please...........

[klark0]

We don't even know whether that patent is actually in use.

[aristotle]

1. Bounce rate up from 48% to 75%
2. Visitor time on site down from 3.48 to 0.35
3. Spam increase and Zombie Traffic
4. Returning Visitors down by 30%

This is probably due to poor targeting. You're getting traffic mostly from search terms that don't match the content of your website. Maybe it will take time for the algorithm to determine what the best matches are.

[timwilliams]

In adwords this is supposedly dealt with by normalizing the data, the stats are only compared apples to apples, in this case your site's stats would only be compared to other sites that have been in the same spot.

[tedster]

We don't even know whether that patent is actually in use.

I think we have a good idea that it is in use. Members here have been reporting this kind of randomized ranking movement for many months, always following website changes. Widespread title changes were the first area I recall being reported.

[Bewenched]

1. Bounce rate up from 48% to 75%
2. Visitor time on site down from 3.48 to 0.35
3. Spam increase and Zombie Traffic
4. Returning Visitors down by 30%

Darn near what we've seen.

1. Bounce rate up from 38% to 49%
2. Visitor time on site down from 2.49 to 2:13
3. Spam increase and Zombie Traffic
4. Visitors down by 30% or more. Returning visitors are actually up a bit.

Since panda/penguin hit we've been working on giving more detailed product information, lessened the links in the navigation.

We've just recently added microdata to help search engines find/display product information properly and changed the size of our image preview from 72x72 up to 200x200 which they click on and get the full size image. We've added better breadcrumbs.

Note: about 3 months before Panda struck we changed the way our large image was displayed, it used to open a pop-up window with it and now we use lightbox.

We use sitemaps for both Google and Bing
We use Google products as well as Bing shopping. (fyi we're now paying for google product listings)

All the things we've done have been to give users better experience with exception of the microdata.

[potentialgeek]

A site in my main site's sector recovered from Panda after selling itself to a "brand."

None of the content has changed that I can see. All the unholy thin pages are now holy.

As soon as I get done posting this, I'm going to stamp an "Inc" on every web page. Lol.

Seriously, do I spend 10K on rebuilding my site to Panda specs or Vince specs?

[realmaverick]

Lower rank doesn't equal lower quality traffic.

I have a panda hit site, due to a simple CMS error. Despite half the traffic, stats remained similar and not only that, with a bit of TLC I managed to almost double the time on site and half bounce rate to name a few.

The site has only been "fixed" for just over a month and hasn't yet bounced back but I'm confident it will.

The only real reason people dont bounce back from Panda is because they don't actually determine and address the true cause(s).

[Andem]

Maybe it's their new patent

Which new patent? I haven't heard of any new patent. If you mean the old patent that was recently hyped up by some Zdnet "journalist", then you might be on to something. The changing of titles and other page elements has been reported over the past few years as a cause of deteriorating SERP rankings.

Bounce rate up from 48% to 75%

It looks like you're using Google Analytics to analyze your traffic and a 48% bounce rate could always mean that you had a doorway page which contained a few words and then a big "click more" link which caused many users to click for more. GA bounce rates don't take time on the entry page into consideration if it's the only page they visit.

The point is the lower the serp ranking the worse the stats are regardless of the quality of the site.

Your point is very well taken. The more frustrated users are when trying to find something (ie. having to dig deeper), the lower they'll dig and the more likely they are to bounce from a page they found in the SERPs. It's no wonder the Adwords results seem to be better targeted towards the query.

Have you ever asked yourself why Google's profits have increased by an obscene amount since the IPO and subsequently more so after Page took over?

If your bounce rate has dropped by such a huge percentage, I believe that: a) whatever you're providing makes Google believe it can better serve their own interests in the form of Adwords; and b) whoever is visiting your site just isn't interested in what you're providing; you're getting the leftovers who haven't been satisfied by the [better] relevance of Adwords results [which are hardly distinguishable].

[indyank]

A site in my main site's sector recovered from Panda after selling itself to a "brand."

None of the content has changed that I can see. All the unholy thin pages are now holy.

As soon as I get done posting this, I'm going to stamp an "Inc" on every web page. Lol.

Have they integrated your site's pages into their domain or they continue to run it as a separate domain, exactly like how they bought it from you?

If it is the latter, it is yet another interesting panda recovery story!

[indyank]

I have a panda hit site, due to a simple CMS error.

I haven't heard that before.Yes, CMS errors might sometimes directly or indirectly lower the search traffic but I have not heard of a panda hit due to CMS errors. Is it possible to throw more light on the kind of CMS errors you are talking about?

[FranticFish]

A site in my main site's sector recovered from Panda after selling itself to a "brand." None of the content has changed that I can see. All the unholy thin pages are now holy.

Have they acquired any new IBLS as a result of being part of a brand family? Imagine what just one link from a big (and related) brand would do to your trust factor.

[arikgub]

Once your down the serps you naturally get abysmal results and algorithmically is impossible to tell the quality of the site.

This might be true in your specific case, but I don't see why it is "natural". I think that what's natural is quite the opposite. When you are ranked at the top, you are much more likely to get uninterested visitors who just click on whatever see first in SERPs.

[creative craig]

... and algorithmically it's impossible to tell the quality of the site.

I think Google and Bing would disagree on this point

[zeus]

I changed my site be cause I was hit in april 2011 by Panda, slowly changed the site, done 2 Month ago, but no change in ranking, my stats is like this
1. Bounce rate down from 82% to 46%
2. Visitor time on site down from 0.45 to 1.55
3. Spam increase and Zombie Traffic : i dont know
4. Returning Visitors down by : I think the same as before

I will not say my site is better be cause of the bounce rate, be cause before they got directly to what they want now that last page has gone, so they get to a page with more options. The cms, design, text... all changed title somewhat the same.

I think Panda just dont work as a ranking factor, the serps are simply pure and in my example, how can a site with 60.000 visits a day suddenly be crap, its no affiliate site or article site.

[muzza64]

I think what seoskunk is seeing is more to do with the fact so many changes were made to the site - if titles and content are changed but traffic is still coming from the same keywords, those pages may not still be relevant, hence the higher bounce, etc. That won't change until Google has fully assimilated the changes, worked out which terms the site should now be ranking for and testing it for those terms (long process). It depends on the scale of the changes.

As realmaverick says, the only reason people struggle to recover from Panda is they are unable to identify the actual problem (and then know what is required to address it).

The example of the site recovering when it was bought by a brand is interesting - just the association with a well known brand would make customers trust a site more which would have a positive impact on user metrics (without any changes except the promotion of the new association across the site).

It's something I've been considering for a while as a way to help my Panda hit site. I believe my site is fairly equal to our competitors (we're working on making it better than the rest) but we're not so well known. I think we can recover by producing a better site, but being associated with a big brand could really improve our metrics and help the site recover. Interesting to hear it has worked for someone.

[seoskunk]

Update - I changed the titles back and hey presto, instant recovery, and my stats went back as they were before

[zeus]

seoskunk - really what is this title thing, a new patent and why does it matter when a webmaster changes a title.

[sid786]

seoskunk - Recovered by changing titles? Awesome! Glad to hear your site's back on top, but what kind of traffic are we talking about here?

[jimbeetle]

So it was just lousy page titles, huh?

[indyank]

He made changes to titles and others because he was hit by panda. His user metrics wend bad in response to those changes. He reverted back the titles and he recovered! I guess he is probably saying his user metrics has recovered and I don't think his traffic has. Correct me if I am wrong.

Now how does your current user metrics compare to your pre-panda metrics?

[seoskunk]

So it was just lousy page titles, huh?

Yep, that and maybe the fact I changed them globally all at once

Now how does your current user metrics compare to your pre-panda metrics?

Traffic is back to before and quality is good so fingers crossed I hoping the site will be OK now

[jimbeetle]

He made changes to titles and others because he was hit by panda. His user metrics wend bad in response to those changes. He reverted back the titles and he recovered!

Yeah, but Panda is only run every so often. The last update was August 20th, so if the recovery was any time after that it's (probably) unlikely that the cause was, in fact, Panda.

In this industry it's very easy to blame something on this buzzword or that buzzword when in fact we should be looking at the basics.

[Sgt_Kickaxe]

Why sites can't recover: They are imposed a glass ceiling on potential traffic based on their aggregate site rating. This way Google can heavily feature the most well known "brands" while rotating everyone else into the mix for the leftovers.

[tedster]

I've worked on site that had a very different result, rather than a glass ceiling of any kind. They were Pandalyzed on the first Panda algorithm release. They recovered during the first Panda data refresh in January 2012. They have now are not simply recovered, they are enjoying record, well-targeted traffic.

Because of this experience I look at many reported recoveries as only partial recoveries. And though we understand more about Panda than we do about Penguin, Panda is still a black box to a large degree. This makes partial recovery a much more common effect than full recovery.

I assume the only reason the site I worked on had such a strong recovery is that their Panda problems essentially came from one major factor that was fixed, and possibly a couple minor ones that were also fixed. It's probably more common for sites to be hitting many Panda triggers in a more major way.

[zeus]

tedster - we need details, what was the major thing :)

[BeeDeeDubbleU]

If he has the holy grail perhaps he wants to keep it? :)

[tedster]

No - it was not "the" holy grail for Panda but it was the important factor for this site. I've written about this in several other threads - the problem was that legitimate republishers as well as some scrapers were getting credit as the original authors. That made the real content owner look like a scraper or re-publisher!

So the goal was to re-establish the site as the content owner and true authority. Since they have about a dozen top-shelf authors with long established industry reputations, that was a very reasonable goal.

We instituted pubsubhubbub with fat pings for every new article; we instituted author mark-up with good author pages for each of the writers; we set a one hour delay on adding the new content to the RSS feed. That's what made the difference for us on the very next Panda refresh.

Earlier on we also fixed some duplicate URL issues, made some successful DMCA requests, upgraded the usability of the site by removing a lot of unused internal linking, and so on. But those earlier changes did not move the needle very much. The big boost came after the steps I mentioned above.

A few new backlinks then popped up from VERY trusted authorities, such as BBB, a security certificates business, and some articles by academic and governmental authorities - but that came AFTER the Panda problem disappeared and gave the site another nice boost. Yes, it appears that good backlinks still can mean a lot ;) But I doubt they would have seen or trusted the site at all before the Panda issue was fixed.

[zeus]

dublicated content, but Matt said that Panda has nothing to do with dublicated content.

[StoutFiles]

Why sites can't recover: They are imposed a glass ceiling on potential traffic based on their aggregate site rating. This way Google can heavily feature the most well known "brands" while rotating everyone else into the mix for the leftovers.

Bingo. Panda isn't broken, it's just pushing all the small sites out of the picture, which is exactly what Google wants. If you compete with a large brand and depend on Google for traffic, you are in trouble.

I've worked on site that had a very different result, rather than a glass ceiling of any kind. They were Pandalyzed on the first Panda algorithm release. They recovered during the first Panda data refresh in January 2012. They have now are not simply recovered, they are enjoying record, well-targeted traffic.

I find these examples to be outliers instead of the norm. The majority of small sites will continue to suffer.