Minor Panda update Oct 13 - per Matt Cutts

[sid786]

Matt Cutts responds on Twitter that they have applied a minor algorithmic update last late night.

Link: http://twitter.com/#!/mattcutts/status/124905069748559872 [twitter.com]

Have your websites been affected with this update? My site's traffic is stable, but I was expecting a positive bump.

[Mod's note: Fixed link so it would display, as the WebmasterWorld link redirect script will break it in most browsers. Copy and paste url into your browser if hyperlink doesn't go to Matt's tweet.]

[edited by: Robert_Charlton at 9:54 pm (utc) on Oct 14, 2011]
[edit reason] fixed link display [/edit]

[MrSavage]

@tedster, I've always been interested in the estimates of affected sites. Cutts has said something similar. Just curious how or where these number are calculated or drawn from. Any clarification on that number would be helpful I think.

It seems more appropriate in my observations to suggest that it only affects those sites that may have been on page 1. What percent of sites were on page one? 15%? Is it to say, page 1 or possible page 2 is reserved now for brand, high powered, big budget sites for searches that incorporate "important" keywords? It's almost what I think is happening.

[tedster]

The estimates that I shared came from private data: a weekly historical record of 100,000's of higher volume SERPs that are then used to calculate the "search presence" for the various domains involved.

Any data source will have various biases buried in the methodology. However, I've compared notes with a few sources who have access to large amounts of data and they're all in roughly the same ballpark. And it does line up with what Amit Singhal and Matt Cutts were sharing. Of course, they have most complete data.

It was interesting on the various Panda iterations to compare what Searchmetrics and Systrix each saw happening. The top level data was pretty much in agreement, but when you got down to which site gained or lost how much, then their data diverged. And of course, "share of search" is not the same as traffic from search. It's not even exactly the same as "search impressions", although that is closer.

Yes, it does appear that more big brands are now ranking well (but not exclusively big brands.) Given what Google said Panda is trying to measure, that's not really a surprise.

Big brands in general are not agile enough to play the technical SEO game and "chase the algo" - but they often can throw a lot of resources (financial and human) at developing good content, analyzing their market and keeping their visitors happy. So I'm not surprised that Panda, on the whole, seems to favor brands. But every iteration has also brought pain to the occasional brand, too. For example, The Motor Report and The Today Show were both losers.

[MrSavage]

Thanks for those details tedster. With perspective, I really do wonder why there are many comments here about poor search results or bizarre results. If people are actually switching to Bing is it really just an issue 15% of the time? Perhaps I'm looking at this wrong. I'm finding the results less useful a lot of time and using a different search engine isn't something I'm doing out of spite. In that regard, if the algo changes were as minor at 15% or so, I can't understand why it's affecting my satisfaction on sites being ranked.

[tedster]

15% of sites being affected would not necessarily mean 15% of searches being affected - the math doesn't hold up that way, because Pandalyzed sites often lost rankings for a variety of query phrases.

[SEOPTI]

It's so funny to read all these nonsense speculations. People should post in the Bing forum instead of trying to achieve time wasting goog rankings all the time (don't be a lemming). It's so boring and a waste of time and energy.

Wake up or get a real job!

Hint: Bing is consistent.

I suggest to invest more energy in web application development than goog rankings. Invent something new and stop thinking you should do this and that and this and that to please the goog gods. You will ruin your life if this is your main income. I almost did ...

[bumpski]

It's so funny to read all these nonsense speculations. ....

The title if this forum is:
"Google SEO News and Discussion"

The only posts that would not be speculation are from Google itself.

And they don't post much here!

[MrSavage]

I can't find the quote from Cutts but I believe he said or used the words "would affect about 12% of search inquiries". I don't remember hearing him say if affecting 12% of "websites". I'm just wondering if the current situation is a little bit more significant that a 12 or 15% affect on searches or websites. This is a small community I realize but our comments about the quality of results is probably enough of a sampling to draw a wider conclusion. I'm merely trying to understand the numbers out there and why I should believe them at all. For all anyone knows, it could be just the people making decent money in a more competitive keyword that are the target. The upper echelon of keywords isn't a number like 25%. It still feels like an extermination process.

[edited by: MrSavage at 3:45 pm (utc) on Oct 29, 2011]

[Reno]

is it really just an issue 15%

Numbers like this remind me of a great quote:

"There are three kinds of lies: lies, damned lies, and statistics."

If 85% of websites are unaffected, there's a logical reason for that: A great majority of those sites never ranked to begin with, so of course they have not been Pandalized. That being the case, we don't really know how many HAVE been deep-sixed post-Panda because they weren't getting any significant Google traffic in any case.

Tedster is exactly right when he says "Any data source will have various biases buried in the methodology." And when much of the data comes from Google itself and their PR man Matt, we have every right to be suspect, which reminds me of another great quote:

"Then there is the man who drowned crossing a stream with an average depth of six inches."

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

[Hissingsid]

If 15% went down 15% must have gone up. I make that 30% affected just half of them will not be complaining about it.

Sid

[indyank]

Your math is terribly wrong Mr. sid, it could even be that for 15% which went down, only 1% of the sites went up...anything is possible...

[Hissingsid]

My Maths is perfect. There's always 10 top ten slots. If 2 of them are vacated they are replaced by 2 other pages.

[bumpski]

I make that 30% affected just half of them will not be complaining about it.

If ten small sites go down in traffic 50% and each had 1000 Pageviews/day and one large site with a million pageviews per day gains that traffic....

The large site will not even notice the change in the first place, and that is why the large site is not here posting "my traffic went up 50% Yippee"!

I know in my case for the site in question I was tying or beating ehow.com (a pretty large site) in Google,s ranking for several terms (actually quite a few terms and topics). Now this site is on page two of the results losing say 10 positions. EHow.com would never notice this new traffic!. It's just noise to them.

Regarding Bing and other search engines:
If your already doing well ranking wise with other search engines like Bing, why would you work on Bing. In this case Google is the only "outlier"

[Hissingsid]

In my niche CTR at #1 for the most important terms is four times what it is at #3. My site has dropped from #1 for two of the most important terms to #3 and #5. This is very bad news for me but the site at #1, #2 etc are perfectly good for Google users. Google can't lose and users probably don't even notice, only those sites that deserved to be at #1 previously lose out unless we can figure out what to do to get back.

[coosblues]

Anyone noticing a small shift in the SERP's you follow. I've noticed what appears to be another small update in the sector I follow. It appears to have happened either late yesterday or sometime today. I'm slowly recovering. I had made a comment when Panda was going on that I had not been affected, but the minor update on the 13 saw me loose a number of spots.

Today, my site picked up half of what it had lost. I spent hours going over each page (all my pages are static html and are content driven).

Here's what I've done so far:

Checked every page for small things like spelling and grammar errors. Double checked all links were working. I did find about 5 outgoing links (on my link page which I never pay attention to) and removed them all.

Changed a few titles to better reflect the content.

I use an HTML editor which creates a sitemap. Along with the sitemap it points out what pages are 404 and why. I fixed every 404 issue and made sure I had a custom 404 page to redirect to my home page.

I had been using the Google plus 1 feature but removed it after seeing how long it took to load, but last night I put it back (I find it odd that the day I put it back was followed by a slight SERP increase)

I had a few sites which were very similar with my main website. I removed any interlinking and even took down one site (just a free blogger site, but again similar content). I had one more site in question ( a very new site) so I just took it down (for now).

Went to my adsense account and changed a number of ad formats/color schemes and reduced how many ads appeared on each page by 1.

I really knew going into this that being proactive might cause me to dip even lower, but sitting around and doing nothing was not an option for me.

I did a few other things that I'll keep to myself, but the above comments were my first proactive steps.

[rlange]

SEOPTI wrote:

Hint: Bing is consistent.

Historical amnesia and short-term thinking. Google was "consistent" at one point, too. Bing won't be consistent forever.

(Also, if Bing truly is consistent, it's only because there aren't widespread attempts to game their algorithm. If Bing aquires a significant majority of the search market, we'll see just how consistent they are.)

Hissingsid wrote:
If 15% went down 15% must have gone up. I make that 30% affected just half of them will not be complaining about it.

Not necessarily; it's likely not a 1:1 ratio. If only one site drops in rank from #1 to #10, then nine sites go up in rank. Total ranks lost vs total ranks gained will be 1:1, but not sites affected.

--
Ryan

[Reno]

Google was "consistent" at one point, too. Bing won't be consistent forever

I think most of us understand and accept that there will be some shifting in the SERPs, on a continual basis. Sites improve so they move up, and others move down ~ that's to be expected and makes for a healthy & dynamic web.

The problem is more serious, where you're going along on the first page for months at a time, in various positions, then one day you wake up and you're on page 4. You did nothing to your site to justify that massive drop, and 30 other sites did not go through a miraculous overnight improvement ~ it just happens, and you know you're scr@wed.. And Google gives you ZERO reason one way or the other for this drop, so suddenly you watch your income cut to shreds, for no discernible reason.

Ruining lives without explanation is unacceptable behaviour on Google's part (or any other public corporation). It's like being stopped on the highway and put in a cop car, and you ask "what's going on?" and they say nothing. THAT is Google, and it's shameful. I'll go further than that ~ it is evil, and the fact that they embrace this FUD policy with such glee says everything we need to know about these people.

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

[Bill_H]

Would it make more sense for this thread to analyze the most recent Panda iterations and what they mean, rather than a semantically focused discussion or vent session? FWIW, aside from short term strong flux in traffic that eventually recovers, we have seen a drop in long tail keywords, while seeing an improvement in most all of our primary keywords. Is that the general consensus of the most recent Panda iterations? And if so, any thoughts on the significance?

Cheers,
Bill

Cheers,
Bill

[Reno]

analyze the most recent Panda iterations and what they mean

I certainly applaud such an analysis, though I hasten to add that I review this forum every day and, to the best of my knowledge (yeah, I might have missed something), there has not been one single clue as to what is going on with Panda. For every person who says they were knocked down for longtails, another will say there has been no shift whatsoever. Others will remove "thin" content with no improvement; still others will cleanup their link structure and will go DOWN rather than up. And why is that? It's as I said, Google offers no clues, no explanation, nothing, just silence. It's like using your credit card at a local store, only to have the clerk tell you it won't go through. So you call the bank and they say your credit has dropped, and you say "how can that be?", and they hang up. That is Google.

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

[AnthonysItalianFood]

@reno ... perfect! You have summed up this mess perfectly! :-)

[sundaridevi]

The problem is more serious, where you're going along on the first page for months at a time, in various positions, then one day you wake up and you're on page 4.

Agree with this. It would be more prudent on Google's part to sort of gradually drop you down over a period of time so that you could realize that something is up and start to work on fixing it before you go from 60 to 0.

Ruining lives without explanation is unacceptable behaviour on Google's part

Yeah, but we must assume that no corporation is going to act in our best interests. About 10 years ago i was emailing back and forth with a former colleague who was bragging about how her online marketing for a software company was working so well ... and all their sales came from AdWords PPC. My reply was, if 100% of my revenue came from google i would be terrified (adwords, organic, whatever). I already get 60% of my traffic from them and it scares me to death. What if they just turn me off one day. She didn't reply at all to that.

I constantly try to diversify my traffic sources. Maybe it helps in that when Panda hit my sites I only lost 25% of traffic rather than 80% like some people describe. But it's always your job to plan for a rainy day ... or an unreliable partner...Sure online as far as search goes, it's not easy to rid yourself of google, but it still is a viable objective. I think the end result is that web developers are going to learn they can't create a world that is spelled google, google, google and expect to comfortably thrive in it. In the end google is forcing us to give Bing, Facebook and whatever else we can find, a much greater chunk of our time and effort.

Back on the topic. I posted in another thread here [webmasterworld.com] that one of my sites lost a lot of traffic over the last two weeks and it seems that a big drop came on Oct 13-14. So I assume it's Panda 2.5. Since I have 4 other sites that have many SEO things done in the same way, and only one got hit, I have a pretty good test bed to see what's really going on.

The research I've done indicates, as before, that Panda is all about onsite factors. It's been written on some other SEO sites that Panda 2.5 is about meta tags. Specifically, if what is in the meta is not on the page, you will have problems. So that may explain the losing or finding of long-tail keywords.

The only other factor that applies to my Panda 2.5 affected site that is not present on my other unaffected sites, is about 10 times more article marketing back links. Specifically, all of the affected keywords are used in article marketing "all in link text". So in this case for my 2 pages that took a -50 hit, we're talking about maybe 200 back links that all use 3-4 word combinations of the same 4 keywords. But those same pages are also in the most competitive niche. I'm really sceptical about whether this alone could cause a penalty, because if that were the case, I could downgrade anybody's site by writing 5-10 articles that all use the same link text, get them syndicated across 10-20 sites each, then watch the guy fall.

Does anybody else have similar experiences or thoughts on what onsite factors are targeted in this update?

[edited by: sundaridevi at 7:44 pm (utc) on Oct 31, 2011]

[Reno]

After my post I was thinking about what I said and want to modify it slightly. If there is ONE thing I do believe has a huge roll in this Panda fiasco, it is the relationship of the site to affiliate marketing. I earned part of my income from my association with merchants, as an affiliate, and I honestly believe Google is looking ultra critically at such relationships.

What I would say is crucial is this:

What is the raison d'être of the website?

I readily admit my sites were constucted ~ well constructed IMO ~ to help me generate income. I undertook this with great seriousness and worked hundreds & hundreds of hours, but in the end, the reason for their existence is not because I have a deeply held lifelong passion that borders on religious infatuation. Rather, it was to make some money.

And to clarify further, I'm not talking about thin MFA ~ I had a ton of extra content. But it may be, to Google, that we & others fall into a broad category that includes the typical MFA, though in my case I did my darnest to separate myself out from those type of websites.

But having said that, perhaps there is no way to separate oneself out ~ if you're in that basket, then all apples are rotten.

So if that's the reason that I and others have dropped, I want Google to just come out and say it in no uncertain terms: "We look askance at sites constructed for the purpose of generating affilate income".

At one time that was perfectly acceptable (for years), but if things have changed, then say so. I'm a big boy, I'll get over it, just tell me what's going on ~ that is not asking too much.

BTW, I have never used AdSense so if I'm right that my own rankings drop is in large part due to affiliate relationships, it is in no way related to AdSense.

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

[bumpski]

there has not been one single clue as to what is going on with Panda

As I've mention before:

Time on Page

versus page size, site size, site traffic. Ratios involving these are being "tweaked"

On my one site that got hit, ONLY, on Oct 13th. The pages that dropped in ranking from 1-3 to 10-13 have virtually no inbound links, and are not highly optimized. BUT these pages were ranking with, or out ranking, pages from vastly larger sites, with vastly larger traffic, with far more inbound links. Why? I think the only thing that could make these pages rank so well for so long was dwell time. Of course the content is very good and exceeds the competitors quality, but frankly, I still don't think Google can measure content quality.

I believe "Time on Page" or "Dwell Time" has been re-weighted with respect to overall page size, site size, and traffic. At least after Oct 13th.

[Lenny2]

Hit on the first Panda... have spent 10's of Thousands "fixing" - no ROI yet. Though, I think we'll figure it out some day. We aren't done getting "free fish" from Google... yet. Traffic over the last couple days has been screwy though lets not forget it's Halloween and Traffic during holidays is always different.

Here is what we've done:

1. Improved speed 50% (we now have THE fastest site in our niche.
a. Removed FB and Google plusone etc.
2. Improved stickiness:
a. Visitors page views by 15%
b. Time on site: 8%

As I said we have not seen any panda LOVE yet... just Panda kicking our @$$. I'm reading lately people talking a lot about looking for "over-optimization" on the page... and removing any traces of being "over-optimized," I'd agree with looking for signs of over-optimization and fixing those things.

Think about it like a social situation. If you are cool then you are the trend-setter... not the guys/gal who walks to the same drumbeat of 1M others. Google, I think, is looking to diversify it's search results by seeking out the trendsetters, who are actively promoting unique content in a unique way. So Panda is, in-part, seeking to qualify the different sites for their unique features.

So for example: if you are an ecom site... selling the same stuff that can be found on Amazon.com... Panda knows that Amazon is the powerhouse of ecom... (a trusted brand that makes it UBER-easy to buy stuff online...) What reason could you possibly give Google to display your ad versus Amazon? I think we'll find that eventually the top 10 results will be the sites that offer the differentiation.

Moving forward the old guard, the "over-optimized-homogeneous-web" is dead.

On another note... It's widely believed that Google is also going to be matching users with websites that fit their user profile. IE. a high-end customer who always shops high end websites will see in their serps a tendency for the high end sites to display first... Probably another reason why Google PR keeps trumpeting the importance of "brand" development for your ecom site. Whether you are high end/low end/ fast at shipping or whatever, Google wants to be able to pigeon-hole (without the negative connotation) you to cater their results to the user.

[sundaridevi]

but frankly, I still don't think Google can measure content quality

Sure they can. Although what you think is important may not be what google thinks is important. Did you ever read the book, "Zen and the Art of Motorcycle Maintenance"? One of the key concepts was, What is Good and What is not Good, nobody needs to tell me this. The point being that everybody pretty much will agree on what is not good. So google can start there. What's not good?

1. Sites where you have to click 5 times to find anything
2. Sites where you have to give your cc number but they have no SSL
3. Sites don't give you enough information to do whatever you came for.


That's just off the top of my head. Can google measure these? Sure as shootin'

1. How many levels deep is it from the home page
2. Does it have SSL pages
3. Does the page have less than x words of text.

Those examples are pretty much my interpretation of some of the things that come out of the 25 post-panda quality questions that Google released.

The question remains however, what new factors have come into play to hit sites that survived all panda updates and finally got hit in this round. One thing that many people who say that were unaffected until 2.5 seem to have in common is that their sites are very old (say 7-10 years). That also presents some very specific avenues of inquiry.

[seaworthy]

Nice post Lenny, long time reader here but registered so I could make a reply to your post.

Although some authority sites have been hit with the 2.5 update, it does seem that the authority/brand site is the way to go heading into the future.

I guess the degree or scale of differentiating individual search results will largely depend on the amount of people who have Google accounts.

I just cannot see Google personalizing my search results if I'm not affiliated with them in any way.

I might buy from Amazon 3 times in a row but the next time they might not have the product I want or I may be still sitting on the fence about a product and need to find a more in-depth assessment. As good as Amazon and other ecom sites are, they don't provide that personal touch that the average joe would reviewing a product.

I understand that Amazon has reviews and such but some people want to see testimonials that are more than just a block of text. They want to see photos of every damn inch of the thing and price comparisons, manuals, etc.

Obviously people are more likely to buy something based on whether other people are as opposed to whether the #1 listing is an amazon site.

Anyway, I have a couple of 'authority' sites that I am developing in advance of the inevitable changes that will occur next year and beyond. In theory large sites would be less dependent upon search engine traffic in the event that they were slapped (which would be rare of course), instead relying on a large community/social base to keep the site from flat-lining.

Ironically it was my big money making sites that took the hit 2 weeks ago. Only now is one starting to return to traffic levels of early October (although it is extremely long tail low converting traffic - I have seen no movement on my core terms).

Some of my smaller, lower quality content sites have improved immensely to counteract that loss of income, though perhaps only in the order of 30% or so. Not much but I'll take what I can get at this stage.

[topstar]

I survived all previous Pandas but not this one. Two of my sites got lower rankings although there are still some first page positions. The sites are several years old and had been stable until now. I think I have been lucky not having to read about and follow Google updates at all this year..

I remember back in 2004 when a site of mine got hit. I then went on to redesign it and it helped a little. So things haven't really changed since then. For me personally I no longer consider keeping websites as a job - it's a hobby and hobbies take time and money. It's fun to build a website people like and reputable websites pick as their "link of the week". If G doesn't like it then there's not much I can do about it :/

Do you have any tips on what to do now? I'm guessing this is all about onsite factors, right? Rewrite meta descriptions, add more text to the main page, a complete redesign?

[netmeg]

What is the raison d'être of the website?

This is an important point.

Despite some of the overblown commentary I've seen across the various blogs and forums, I don't think Google has necessarily declared war on all affiliate marketing. (I have a couple aff sites and I know people who have many more (and better) sites than mine who have never been hit) Google isn't trying to get into every marketplace - they couldn't if they wanted to.

To me, a lot of this Panda stuff seems to go back to the very foundation - the business model. What are you doing that's better and different than everyone else that's doing the same thing? Sure you can add tons of content (and value) to your affiliate site, but if it's the same type of content (and value) that many of the other affiliates offer, it's still not good enough to rise to the top - even though it's a great site. If Affiliate A has a page with videos and tips on the care and maintenance of his widgets, and Affiliate B has a page with a photo gallery and FAQ on the care and maintenance of his widgets - the pages may look entirely different and be worded entirely different, but at some level, it's still the same thing. But then Affiliate C comes along and puts in some kind of a discussion / group plugin so that his users can interact with each other about how they take care of their widgets, and suddenly Affiliate C is offering just a little bit more.

I love affiliate marketing as much as the next person (and I wish I were better at it) but the inherent problem with it is that there are always going to be a bunch of people doing essentially the same thing in essentially the same way, and they can't all be the authority. So you have to come up with something the others don't. And it has to have real value to your users - not just something you tack on as an afterthought. And as soon as you do, some of your competitors will copy you and you'll have to come up with something else. And you have to find ways to make that difference very clear to your users and very clear to Google as well.

The same is pretty much true of ecommerce as well, although if you're a stocking merchant or a manufacturer of product, you may have a little better chance of standing out if your product itself is unique. If it isn't, then all the above applies.

And this is also how you compete with a big brand.

FWIW this is all pretty much the way it works in the offline world too. It's hard.

[tedster]

Do you have any tips on what to do now? I'm guessing this is all about onsite factors, right? Rewrite meta descriptions, add more text to the main page, a complete redesign?

Welcome to the forums, topstar. No one outside Google knows for certain what Panda is "all about", but it's safe to say it is a lot more than on-page factors. There have been reams of opinion and observation written about Panda this year but nothing is conclusive.

There seems to be a lot of traffic data and "social mention" metrics folded in - but the overall Panda algo is a complex bit of machine learning and I assume that the AIM is what Google said it is. That aim was a metric for the perceived "quality" of the content (as opposed to its backlinks or relevance. However, that statement is so vague that we're pretty much left with Matt Cutts' advice to "chase your visitors, not the algorithm."

If you want to dive into things a bit more, here are a good couple threads about Google's official communications:

Matt Cutts and Amit Singhal Share Insider Detail on Panda Update [webmasterworld.com]
Quality According to Google - Official "Guidance" on Panda Update [webmasterworld.com]

[sundaridevi]

@topstar: I have 5 sites, all which have quite a few top 1-10 rankings and improved rankings in every Panda update until Panda 2.5. In the Panda 2.5 update, one site had some of the most popular pages significantly downgraded. Very little is different between the one site and the other 4, and very much is the same. As a result, I am looking into very specific areas for change. I have some specific ideas, but since in this forum whenever I post specifics or requests for specific I get no answer. I'm sure some people know more than the general level of participation here suggests so I'm looking for others who were unhit until Panda 2.5 to exchange data with in order to confirm some of my ideas. PM me if you like.

[topstar]

I'm now reading more into this and making modifications, starting with noindexing low quality pages. Two sites that were affected had 2 to 5 position drops for their main keywords. It doesn't sound much but the date October 14th matches so I think it was caused by the same algo tweak that has caused more significant drops for other people. It might be an automatic/manual devaluing of some backlinks too. Are there any indications that some manual penalties were also imposed on that day?

As for quality one of the sites that overtook me contains nothing but links and hasn't been updated for several years. Authority status and high quality backlinks really help a lot.