Panda Refresh in 2-4 Weeks - Finally after 6+ Months!

[SnowMan68]

After 6+ months we are finally going to see a Panda refresh in the next few weeks.

At SMX Advanced tonight, Google’s Gary Illyes announced that the next Panda update will happen in the upcoming weeks. He said he expects it in the next two to four weeks.

Illyes referred to it multiple times as a data refresh, not an algorithmic change. So sites that have been suffering from this algorithm may see a recovery in the near future. However, not all sites will see a recovery: Some may not recover, and new sites may also be hit by this data refresh.

[searchengineland.com ]

[JS_Harris]

How many sites change drastically, in terms of quality (or even in terms of the quality signals that the algorithm looks at), from day to day or week to week?

For me about one a month. I own just a handful but I make a point to go over them regularly and update them, sometimes to an extreme degree as was the case for my recent move to full mobile and schema compliance which I combined with my new social strategy.

Webmaster tools show a huge spike in number of pages crawled and amount of data collected over a 3 day span after the most recent changes I made so I submit that how often Google refreshes the content is irrelevant because I can force a full refresh just by making changes on-site. What I can't do is force Google to give me Panda credits for those changes, a big bear stands in the way(I've not been Panda smacked, but if I needed to please the bear...)

My pet peeve is that it's the same content and the changes have more to do with pleasing search engines and browser whims of late. I believe there are a good number of sites online getting traffic they don't deserve and many more not getting the traffic they do deserve, I feel that should all be based on content but we're not there yet.

[rish3]

Is there a compelling need for the data that the algorithm looks at to be refreshed constantly or even frequently?

Google isn't pitching that rationalization. The Gary Illyes comment makes it clear it's not refreshed because they have low confidence in the new data. Makes you wonder how much confidence they have in the data that's being currently used.

If there wasn't a compelling need for reasonable refresh, then why would they fund a project specifically to do that? (Everflux)

[EditorialGuy]

Google isn't pitching that rationalization.

It isn't a rationalization, it's a rationale. :-)

If there wasn't a compelling need for reasonable refresh, then why would they fund a project specifically to do that? (Everflux)

If constant refreshes of Panda data were a compelling need, they'd probably rank higher on Google's "to do" list than they do.

There's a difference between "Here's something we need to work on" and "OMG, we can't have people complaining on the Webmaster and SEO forums, so let's shove something out the door right now!"

[seoskunk]

By your own admission this algo creates false positives.....

Of course it's an imperfect algorithm.

Yet knowing this algo creates false positives you advocate not refreshing regularly

So a flawed algo infrequently refreshed determines the SERPS.

[rish3]

There's a difference between "Here's something we need to work on" and "OMG, we can't have people complaining on the Webmaster and SEO forums, so let's shove something out the door right now!"

I wasn't implying it should be a high priority because people are complaining. I was pondering why it was once such a high priority that it was a funded, named, project. Google it if you like: "Panda Everflux".

My guess...they aren't happy with Panda. Think about the sequence...

Panda -> Talk of Everflux -> Release of Kinder/Gentler Panda -> Letting it go 6+ months with no refresh

Kinda looks like what happens when you lose confidence in a product, right? You dial it back, stop investing, etc.

[Whitey]

@glakes: I suspect that both panda and penguin scores are overridden by brand assessments in Google's algorithm. That would explain why many thin pages on branded domains have been ranking so well and small businesses are hard to find on page 1 of Google's serps. As long as Google allows the brand signal to take precedence over actual content quality, many lesser known/small websites will continue to be handicapped in Google's search results

@glakes : Agreed. I'm observing a strongly branded ecommerce site that became an affiliate 4 months ago, using the same content as hundreds of small sites. All the content is thin content - nothing original. It still ranks, probably because of no update occurring.

I wonder what it will do after the update. Rank probably.

GOOGLE PATENT: Determining Site Quality Score [webmasterworld.com]

Brand ranking bias would seem to be the intent behind Google's thinking in it's patent and the planned direction of it's SERP's from way back.

[EditorialGuy]

Yet knowing this algo creates false positives you advocate not refreshing regularly

So a flawed algo infrequently refreshed determines the SERPS.

This thread isn't about algorithm updates. It's about data refreshes. They're two different things.

No one here is suggesting that the Panda algorithm shouldn't be updated whenever improvements are ready for prime time.

[seoskunk]

This thread isn't about algorithm updates. It's about data refreshes.

Nor was my answer, please read more carefully in future.

you advocate not refreshing regularly

[RedBar]

Has this already happened this weekend?

I've posted in the monthly thread that I am seeing improvements for already good SERPs positions for my widgets, something's definitely changed in my neck of the woods.

[EditorialGuy]

you advocate not refreshing regularly

I've advocated nothing of the kind. If Google wants to refresh data in real time, that's fine with me. I've merely suggested that there may not be a compelling need for constant data refreshes (not to be confused with algorithm updates), since very few sites have significant increases or decreases in quality from day to day, week to week, or even month to month.

[rish3]

very few sites have significant increases or decreases in quality from day to day, week to week, or even month to month.

I would suspect that new, high quality pages do get created on Panda afflicted sites, assuming the owner is trying to recover.

I also assume that crappy pages get created on sites that are not currently affected by Panda, but might be if they were re-scored.

Last, brand new sites/domains get created on a regular basis.

6+ months is quite a long time.

[EditorialGuy]

I would suspect that new, high quality pages do get created on Panda afflicted sites, assuming the owner is trying to recover.

Maybe. But we've been told that Panda is a site-wide algorithm, which means that it could take a lot of new high-quality pages to bring up the average (or a lot of new low-quality pages to bring down the average). Just as important, the quality of individual pages is likely to be just one factor in how Panda scores are calculated.

6+ months is quite a long time.

Google does have a reputation for taking the long view. :-)

[Nutterum]

I have a question that was always a dodgy subject when it comes to Panda. It is about the quality of the copy on a page. Does Panda recognise poorly written content? Because I have seen more and more landing pages surging on the serps with copy written by 2nd grader learning English as a third language. I have searched quite a bit but no concrete proof was provided nor official response being confirmed on the subject.

[Johan007]

Does Panda recognise poorly written content? .... I have searched quite a bit but no concrete proof was provided nor official response being confirmed on the subject.

There is no proof but the technology does make this a possible ranking factor (Note grammar is easy and free to check online these days). If you think about it, this is more important than the quality of code.

What I do know is that panda is after unique and useful content. In fact I have removed useful aside content in my articles that help users because the content could be found on other websites! This content is still held in the database so I could re-introduce this information when I have more clarity that it is not doing harm.

[RedBar]

Does Panda recognise poorly written content?

Whether it comes under Panda I am not sure however I am positve that a couple of years ago or so Google said that poorly written pages would be penalised however, like you find, there seems to be an awful lot of garbage ranking and as for keyword stuffing, don't even get me started!

[Nutterum]

I did notice similar pattern before the previous data refresh was introduced in 2014. Several websites appeared out of no where (after some digging they were appearing in Page 3+). All these websites were glaringly written by a single individual,, or a small group of people, with basic grammar and very poor flow of language, in hopes of scoring a good rank on the niche and capitalize as much as possible before the penalty kicks in. All websites had almost 0 backlinks and were keyword stuffed to the point of nausea. After 2-3 months Panda swoop them under the rug, leaving almost no poor quality, websites afloat. Now in 2015 for a little over a month I am seeing those same (and some new) websites re-emerge from the bottom of the SERPs, just in time for Google to announce their next Panda refresh. I am not a believer of coincidence and think, that the same can be said for many SERP vectors and niches.

Have you spotted similar behaviours in the SERPS close to Panda release?

[rish3]

Using spelling or grammar as an indicator of quality, in my opinion, would be a terrible idea.

Consider the tech niche for example. Many of the top experts use a language for published material that's not their native language.

Or, going the other direction, there are ways to create a grammatically stellar content farm...that has nothing new or useful to share. In fact, many of the existing content farms have a submission process that checks for grammar mistakes and rejects the submission until it's in compliance.

I would much rather read content that's flawed from a grammar perspective but rich in terms of subject matter expertise.

[RedBar]

Using spelling or grammar as an indicator of quality, in my opinion, would be a terrible idea.

Try telling to a "foreign language" buyer for one's widget products that you couldn't be bothered to have information about one's products professionally translated and see what reaction you get!

[EditorialGuy]

I would much rather read content that's flawed from a grammar perspective but rich in terms of subject matter expertise.

I'd much rather read content that's rich in terms of subject-matter expertise and written by someone who knows how to write.

There's no reason why it has to be one or the other. If I ran a search engine, I'd say "Want to rank? Give me both."

[Nutterum]

By the way checking my WMT accounts I am seeing massive surge on almost all properties in terms of indexed pages. The update is coming sooner than most think. Last few rounds it came less than a week after similar crawl burst.

[rish3]

There's no reason why it has to be one or the other.

Yes, there are. I mentioned one common reason.

There are other reasons as well. "Grapes of Wrath" is riddled with poor grammar, for example.

[Robert Charlton]

"Grapes of Wrath" is riddled with poor grammar, for example.

Search engines... all of them, not just Google... have problems with certain kinds of language, things like humor, similes and metaphor, idioms, irony, ambiguity, and probably regional dialect. There are questions about grammar vs frequency of usage.

That's a very long discussion, and an in depth discussion is probably off topic for this thread, except to say that language which departs too far from the norm is going to give a machine some trouble with interpretation, and context can become extremely important if you want to rank.

In my experience, Google handles context better than most search engines do, but it is definitely still learning.

[rish3]

That's a very long discussion, and an in depth discussion is probably off topic for this thread.

I agree, but did not want the thread to end with the notion that "good grammar" was some short of shortcut to finding quality content. It isn't. For the two reasons I mentioned, and many, many more.

[Johan007]

In some sectors where content quality is excellent (e.g movie reviews) then you need other factors such as the quality of writing, uniqueness, unique imagery etc to set the pages apart from the rest. I don't think anyone is claiming one factor is a silver bullet.

Still no change in my 2011 Panda affected site after 6 month cleanup process to remove all forms of duplicate content. Not long to wait, I am truly excited.

[Nutterum]

Hence why there are more than one occasion where, synonyms placed like "variety of products" in a different section, often fool Google in to ranking them better and so long you have decent quality on the Page you will have higher chance of not getting Panda-ed. Hell i`ve seen some websites having these sections turned off or on from the back end when an update is approaching and manually requesting the bulk of their websites reindexed, so that as soon as the storm ends they can do the same thing all over again and spam the top 5 SERP positions. I've seen this in most widget verticals and some service sectors. Is this the new anti-Panda grey hat SEO? If so I don`t really like it and unless Google starts distinguishing text better, I don`t see this going away anytime soon.

[MikeNoLastName]

Might finally be Heeeeeeere! According to WMT AND our logs just went to 4X normal crawl rate and 2X highest rate in 6+ months. Over 3X std dev.

[n00b1]

I'm seeing shifting that smells like a panda.

[webmuppet]

Might finally be Heeeeeeere! According to WMT AND our logs just went to 4X normal crawl rate and 2X highest rate in 6+ months. Over 3X std dev.

Is faster crawling a sign of Panda, I've never noticed it in the past. Why would Google need to spider more than normal, surely they already have the data they need to d a refresh from the normal crawling.

[Shai]

We are observing some ranking movement across a wide range of our client base. Some more violent than others. I am, as of yet, not found a common denominator for the most unsettles sites. I think it may be a tad too early to point fingers at any Pandas just yet but with the recent chatter its a strong possibility.

[Johan007]

I think it may be a tad too early to point fingers at any Pandas just yet

Agree. Crawling has become more aggressive but naturally everyone is requesting and submitting new content via WMT fetch.

Hell i`ve seen some websites having these sections turned off

I did this last week. I have some useful aside information that can be found on other sites along side the unique content, but felt I can not risk showing it. I may introduce it back but will assess the risk at a later date.