The problem with Panda analysis, and the solution

[whatson]

All we seem to be doing, in our steps to isolate the factors that cause sites to get Panda'd is comparing what our sites have done to put us in this situation.

I think the best way to develop a solution would be better to compare sites that were not affected, and report on what things we have done, or not done on these sites.

I am fortunate enough to have sites that have not been affected, and some of the major factors of these sites are:

- Absolutely unique content
- No pages with less than about 150 words of content.
- They do have adsense
- Almost all the content is old, and not up to date
- Bounce rates are 10%

higher

than panda's sites in the same industry.
- Contains pages with heavy content, perhaps up to 2000 words on some pages.

That is all I can think of at the moment, but would be interested in what others report about their unaffected sites.

[Shatner]

The problem with this method of analysis is that many of the pandalized sites have the same things.

For instance, your description could have been of my site. Except my site was Pandalized and yours wasn't.

[tedster]

Here's an interesting angle posted in an article on Search Engine Watch. The Panda winners this time around are, in many cases, websites with a lot of video content.

YouTube was the biggest winner, according to Searchmetrics, as were a number of big brand video sites - Hulu, MTV, NBC, CBS, HBO and a number of others that contain a large amount of video content.

[searchenginewatch.com...]
My own take here is that the video angle is a correlation, not a cause-and-effect factor. But if the public is responding even more to video right now, that would make all kinds of metrics for video sites start dancing.

The thing about this analysis at SEW is that it is focused on winners, and not only sites that were not hurt. And one other caveat (as was noted here in another thread) the SearchMetrics data is confined to search presence - rankings across a wide variety of keywords. That may or may not translate into a change in the total search traffic that Google actually sends to the sites.

I have sites with a decent amount of video that have never been Pandalyzed. At the same time, I have sites with NO video that have not been hit, either.

How does video look as a factor for others whose sites escaped any Panda trouble this time?

[mhansen]

I wish it were all that easy, but I think you also need to factor in:

- Niche Market
- Demand Volume

In one of the markets I tend to play in, its now being Owned by MFA sites, where 6-8 months ago, or BP, (Before Panda) there were NO MFA sites in the top 20 serp's. There have been little to no changes with each new Panda rollout.

This leads me to think it may be related to this patent: [webmasterworld.com...] in a lateral sense, to filter low-demand or low-result content out of the Panda algo, until it becomes mainstream at the content farm level.

If its a tight niche market OR a market with low amount of relevant content for query > don't pandalize, otherwise, measure and rank against panda algo.

MH

[Reno]

But if the public is responding even more to video right now, that would make all kinds of metrics for video sites start dancing.

This is a very interesting observation and I thank you for it. Many of us are not in the financial position to shoot/edit/encode/upload quality videos, but perhaps adding a link on many primary pages to video at YT might be seen favorably by Panda. Worth a try when not much else is doing any good....

..........................

[AlyssaS]

whatson - the only way to do a really accurate comparison is to set up test sites - the advantage being that you don't care what happens to them so it is easy to push the limits. You need two test sites based on the same niche, and then vary elements, such as word length and so on and see which does better.

[AussieWebmaster]

this change seems to bring new elements - video, +1s and bounce rates/time on site

while the update is seen as a Panda version but is more than content farm - more popularity of content elements

[tedster]

Many of us are not in the financial position to shoot/edit/encode/upload quality videos

When video started taking off as a clearly important marketing component, I (along with several of my small business clients) had a similar reaction. But you know, as I invested the time to get educated in it, I found it didn't take nearly the full-scale production ability that I feared.

This leads me to think it may be related to this patent: [webmasterworld.com...] in a lateral sense, to filter low-demand or low-result content out of the Panda algo, until it becomes mainstream at the content farm level.

Thanks - that is a solid observation, in my book. Remember how the second iteration of Panda was said to go "further into the long tail"? Why bother running an intensive algo over sites that aren't even in a market niche that has much content in the first place?

[nippi]

YouTube contains a comparatively small amount of video content compared with it's on topic, user submitted, user rated written content. The videos are just link bait and free unique content writing attractors.

[ScubaAddict]

I have several sites in the same niche - education.

Pandalized: 1 large site = 1000's of pages, covering many topics, very popular, WAS in the top 3 search results for hundreds of keywords, several somewhat competitive and the site is mature (13 years old). Pages are content rich, 3 to 5 ads per page depending on length (all unobtrusive - no popups, layers, interstitials, audio, etc). No links ever bought or sold.

Non-pandalized: 5 other sites, much smaller (10-50 pages max) focused on a single specific subject ranked in the top 3 for the specific subject (2-5 keywords each), somewhat immature (3-5 years old). 1 to 2 non-obtrusive ads per page.

No video on any of the sites, nor on any of the sites that currently rank in the top 3. My site is unique in that it has dynamic and innovative useful content - and I feel google's algo can't understand the complex structure and value of the content I produce.

The sites that now rank in the top spots (and have since Feb) are small and very static (compared to mine). Some are very old and never updated, and one that stole my most profitable top spot is simply page with nothing but a single link to another page. The 2nd result is the page the first spot links to. Really Google?

One thing that is almost universal is that the top spots are now occupied by very small immature and mature sites with 1 banner or no advertising. Design is almost universally very poor and unprofessional (think just plain text and links - no images, maybe funky background colors). A few attempt to be dynamic in nature, but are amateur. The others are just static and limited in their value and "replay-ability".

This gives me the impression that my pandalized site covers too many subjects, contains too much content (which is too complex in nature for google's algo to understand), and has ads to support it (we don't sell anything).

We do have a decent Facebook following - about 7K after starting a facebook page a year ago.

Still looking for answers and making changes (completely blind of course).

[netmeg]

No sites (or client sites) have ever been pandalyzed; only two client sites make good use of video. It's on my to do list for some of my sites, but I just haven't gotten to it yet.

[AussieWebmaster]

the spaces I work in have had some interesting changes with this update - there are some real bogus newcomers getting high rankings I have never seen before and who copy much of what the sites that have dropped do

[whatson]

Wow Netmeg, you just be very happy with that. Care to share what is your are/aren't doing to have left you unscathed.

[netmeg]

I have, at various times, in many items in this forum. Largely resulting in derision and name-calling. So, no. I'm done with that.

This last iteration didn't seem to affect any of the sites under my control, but at least two competitors had a lot more keywords show up in the top two pages, according to my favorite keyword tool. Once I get over the monster killer death cold, I'm gonna have to drill down on that.

[AussieWebmaster]

nutmeg - I always listen

[Hissingsid]

@Tedster
My own take here is that the video angle is a correlation, not a cause-and-effect factor.

Video could increase average duration of visit on site. Related videos may encourage exploration and reduce bounce rate. On topic comments keep content fresh. Perhaps balance between outlinks and backlinks right on video sites.

Cheers

HS

[whatson]

I think I have just coined the term for the this RTA - Reduced To Adsense.