Was SEOmoz affected by Google’s Panda Updates? It depends how you look at it. Since the first update hit in February of 2011, organic search traffic to SEOmoz has increased by 49%.
To be fair, the dates that Panda hit don’t match up to periods when we saw traffic gains. In this time period we’ve rolled out original content, popular blog posts, introduced new tools and made several SEO improvements. In a way, you could say we’re good at not getting penalized by Panda.
Looking at the sites that lost the most traffic when Panda hit, I’m amazed at how poorly formatted most of these sites remain, even months later. A few have made improvements, but it feels like many webmasters decided it just wasn’t worth the effort, gave up, or they simply didn’t know what to do.
Panda – The 2-Minute Nutshell Version
Panda starts off with human quality raters who look at hundreds of websites. Computers, using machine learning, are then brought in to mimic the human raters. When the algorithm becomes accurate enough at predicting what the humans scored, it’s then unleashed across millions of sites across the Internet.
The point is: Panda starts off from a human point of view, not a machine’s. We can look at these sites with human eyes and see the obvious.
Remember, Panda is a site-wide penalty, not a page penalty. So if a certain percentage of your pages fall below Panda’s quality algorithm, then the whole site suffers. Fix enough of these pages and you may recover.
Note: I’ve used actual screenshots taken from sites well known to be hit by Panda. I don’t mean to call anybody out and I’m not picking on any particular site. On the contrary, we all have a lot to learn from these examples. Also, a lot of these topics are covered in Dr. Pete’s excellent post, Fat Pandas and Thin Content. It's well work a look.
1. Heavy Template Footprint
Do you ever look at a site and ask yourself, “Where’s the Beef?” Consider the page below. How much original content exists above the fold?
This template footprint creates a low ratio of original content. No one knows the exact threshold for what qualifies as duplicate content from a machine point of view, but this clearly violates all human standards.
Here at SEOmoz, our PRO platform uses a 95% threshold to judge duplicate content. This means if 95% of all the code on your page matches another page, then it’s flagged as a duplicate. To check your own ratios, try this nifty duplicate content tool.
2. Empty Content
Do you see pages that exist simply to link to other pages? I found this high authority page from faq.org in less than 10 seconds, which indicates there are plenty more.
Yep that’s the whole page right there. Or how about this page from Suite101? Eliminating these types of empty content pages, or just adding good material, will go a long way in eliminating Panda penalties.
3. Overlapping and Redundant Articles
Each page of your site should address a specific topic, instead of addressing a slightly different variation of a keyword phrase. How many variations of "Acai Berry Cleanse" does a website need?
The example above is from a high profile "content farm." This is one of the many reasons I believe Panda hit article sites so hard – pages and pages of overlapping articles that targeted keywords instead of humans. Combining these articles into a few, highly usable resources would cut down on the confusion.
4. High Ad Ratio
I understand the temptation. We can’t escape it. Google even tells us to plaster ads all over our site. But the Adwords team is separate from the spam team lead by Matt Cutts. Don’t expect them to both give the same advice.
Optimizing for Adsense does not mean optimizing for search. I’m glad Google keeps their departments separate, but more consistent messaging from the company as a whole would reduce webmaster frustration.
5. Affiliate Links and Auto Generated Content
If a machine built your pages with minimal human intervention, Google wants to devalue you. We see this time and time again with multiple affiliate sites across the web.
Disclaimer: I predict that someday machines will be able to produce web pages indistinguishable from human generated content. Until that time, avoid it.
Beating Panda – Reduce the Sins
It’s my personal belief that the above 5 sins, alone or in combination, account for the vast majority of Panda penalties. Yet webmasters seem perplexed when faced with fixing the these problems.
Dr. Pete pointed me to a post on Webmasterworld from Duane Forrester. Although he’s talking about Bing, he explains the Panda situation well.
If a user searches on Bing, we return the best SERP we can for the query. The user clicks on the first result (as we’d expect, normally). They hit YOUR website, and…
1 – are so engaged they forget about the rest of the Internet for a few minutes and bask in your glory, getting the information they wanted, and more.
2 – are so dismayed they immediately hit their back button, suddenly popping back onto our radar, alerting us to the fact they were displeased with the result we just showed them – why else would they suddenly come back, without consuming your content? Displeased.
-Duane Forrester
This is the future of the web – when the equivalent of human eyes look at every page of your site. Would you want your father to visit your website? Would you want your mother to shop there?
Rand had it right when he told us to improve engagement metrics. Make legitimate, on-site fixes that actually improve your visitor’s experience. You’ll find that reducing bounce rate, increasing page views and time-on-site has the side effect of making your visitors, and you, happier.
Interesting. I'm sure that a lot of sites that have suffered share these "sins".
However, there are plenty of high-end directory sites that have committed at least 4 of these sins and are not suffering at all, particularly if they are well-established.
Machine built web pages are fine, so long as the underlying data content is high in quality, unique and so forth.
Saying that a page "doesn't deserve" to rank unless it has been lovingly hand-crafted by a real person may well be applicable in the case of editorial-based sites - no one wants to read gibberish - but it ignores the very substantial value that data-led sites bring to the web.
You have a good point. Perhaps the future is now. What about the travel sites that I visit to buy my airline tickets? For the most part, they are useful and completely machine generated, mostly with content that can be found on other places on the web. Thanks for the alternative point of view.
@cyrus - add Google to this. A website that is totally generated by machine - using first generation AI
On the flip side, I know of sites that offer extremely valuable and unique content, but because that content is created by users and the site has more on-page ads than average, it's been slapped by Panda pretty hard. Which is a shame because the information is genuinely useful and quite frequently can't be found anywhere else on the web.
Just as a related FYI, here is some info for anyone looking to diagnose severe changes in Google Analytics traffic since August 11 (I have not seen this discussed on SEOmoz yet). GA made a change in the way sessions are counted that was not supposed to severely affect data but has ended up affecting lots of users as can be seen in the comments on their official blog post, comments on the SearchEngineLand post, many posts in the Google Analytics support forum such as this one, and many tweets to @GoogleAnalytics.
So if you think your visits have severely gone up in the last few days while at the same time significantly increasing your bounce rate and destroying your average time on site, it may have nothing to do with Panda or your content. Wait for the dust to settle on this thing first (I hope it's soon.. maybe SEOmoz can shed some light on it.)
I've got additional 5k organic visits in 1 week, but number of unique visitors is the almost same. So there are no problemm to determine real trend. Waiting for multi-channel funnels.
I've got additional 5k organic visits in 1 week, but number of unique visitors is the almost same. So there are no problemm to determine real trend. Waiting for multi-channel funnels.
Hi!
Excellent article, but at one small point I tend to disagree.
You and Dr Pete cite
If a user searches on Bing, we return the best SERP we can for the query. The user clicks on the first result (as we’d expect, normally). They hit YOUR website, and…
2 – are so dismayed they immediately hit their back button, suddenly popping back onto our radar, alerting us to the fact they were displeased with the result we just showed them – why else would they suddenly come back, without consuming your content? Displeased. -Duane Forrester
What about search terms like "How tall is the Eiffel Tower?" If searchers get their question answered immediately and jump back, yout got to watch their beahviour, don't you? Sometimes i need quick answers, and i need to spend only seconds on a page, I would be displeased if I needed longer. That would waste my time.
"What about search terms like "How tall is the Eiffel Tower?" If searchers get their question answered immediately and jump back, yout got to watch their beahviour, don't you? "
I wonder anyone will bother to click on a search result for a single word/line answer. Search your question on Google and you will get the answers via meta descrptions of the search engine listings. No need to click any search result.
Yes. And there's also a difference between clicking on a site, getting your answer, and moving on to other things.... and clicking on a site, not getting your answer, going back to SERPs, and clicking on a different site. I'm guessing (hoping?) that search engines are getting smart enough to tell the difference.
You'd think that, but if people were that clued up then 'facebook' wouldn't be one of the most searched for terms - people would just go straight to the site. I actually watched my friend navigate to Google.com, type 'google maps' into it, and then search - bypassing the search box in her browser and the clearly marked Maps link on Google itself. To say I was scared would be an understatement.
The forum thread goes into detail on that, as you're absolutely right - there are many reasons a visitor can bounce off a site, some of which don't indicate a bad site. Blogs, for example, have very high bounce rates, as people often read one article and leave. It doesn't mean the blog is bad.
Duane clarified that:
(1) This is just one of many signals (he claims Bing uses about 1,000)
(2) They're looking at very rapid bounces (in the milliseconds)
It's not an easy problem to solve. No one metric can adequately encompass quality. That's why Bing uses those 1,000 signals and Google has 200+ ranking factors (many of which are a combination of signals).
Panda represents many variables that may be fed into a single ranking factor of sorts (PandaRank?). So, even if Google is using user signals, they're feeding them into yet another model that's balancing them with other signals.
Well the on-page-time or bounce-back-time is probably compared to other results to that query. If every SERP has on-page-time of 3 seconds on "how tall is the Eiffel tower" then that metric has basically no effect.
I must be penalizing every site on the SERP because I open everything in a new tab so I don't lose the original SERP. Would this affect Duane Forrester's second comment?
You have to see Google AdSense in a different light. Why would Google penalise you for placing THEIR product on your site? It probably is an exception in all fairness. This theory isn't proven by any means however it is logical. Does anybody agree?
Google has to be impartial to some extent. I think that rather than Adsense ads specifically being targeted either way, it's the ratio of ads to content in general, no matter what those ads are or where they come from. In which case, we may well see other advertisers start to promote smaller ad sizes overall. We've already been seeing ads integrated into content a lot more in recent times. It would be interesting to find out if Google detects those infolink text links as advertising or not.
So generally you are agreeing with me, somebody would need to have a very strong excuse not to agree with me really. In terms of it being the ratio of ads to content in general, yes I agree with that however I still feel, how can Google penalise a website generating them money?!?! It doesn't make sense. Let's say a home page had 95% Google AdSense and 5% unique content then they SHOULD get penalised but I really don't think they would because you're promoting Google. I would love to test this however I don't have the time nowadays :-( I'm sure somebody would like to attempt it :) Thanks Jenni @ Heart Internet
Thanks for a great article.
My site was hit badly on the first Panda update. So I Got rid of google adsense banners on the top of the each page, removed some affiliate links(leaving one on each page) and put some articles up which had no adverts or affiliate links on them. I now recieve about 50% more traffic from google than before the update.
Thanks again.
Could you possibly write a YOUmoz post on this study and present some kind of Case Study? I am sure it would be a very interesting read for members of the SEOmoz community.
I agree Mysterio.
I think it would be interesting to see how the affiliate links were weighed to determine if it stayed on the page or was moved. I am guessing that it was determined by its relevance to the content on the page. I would love to know more.
Thanks Cyrus for the great post!
Good post cyrus...Google panda is a trendic topic here in paris and it's good to have these wonderful and simple examples to show to our clients and colleague :)
Hi Cyrus,
Great post. One question though. Of the Duplicate Content parameters messured by the tool you refer to, which of them are the one you look at with the 95% threshold?
Is it HTML distribution value, Total HTML similarity, Total text similarity or something else?
Thanks
I was wondering the same thing
I will 3rd this wondering. The results on the tool vary from close to 95% all the way down to 45%.
One souce of redundancy is structural. In Jive Software's system each Reply to a thread is its own URL to the entire thread. We had to limit Google's access to that directory to try to keep it from thinking we were pushing redundant content. Even so, it has been very difficult to stop the indexing. Very unpredictable reactions from Google.
I would hope it's text similarity.
SEOmoz measures total code, similar to toal HTML similarity. 95% similarity is a good threshold to use as a benchmark, but it's good to use your own judgement as well.
Those deadly content sins are so obviously that I am wondering all the time how those pages could have survived / have been successfull before they have been hit by Pandy.
I guess most of the internet users click the back button if they reach those kind of pages or like you cited Duane Forrester - are "so dismayed" they leave the page at once.
It's amazing that in a few short months Panda has already elevated what we expect from the web. There's still spam, but how soon we forget what it was like before.
Same as Pop-ups, Spammy sites are on their way out.
Go Panda Go!!
Although I must agree that all points mentioned in this post will improve user experience and make a better website (no doubt), I think the SEO community still have to investigate further Google's Panda Algo Update.
I am not convinced ...
I still see sites full of ads (Adwords, whatnot), dupe content or lack of it, ranking as high if not better than ever.
In my opinion, we have not solved this issue. I know of sites which lost traffic with the first Panda Update and recovered recently, having made no changes at all even, so? Google might be simply reverting changes, not fine-tuning Panda as some seem to suggest ...
"I still see sites full of ads (Adwords, whatnot), dupe content or lack of it, ranking as high if not better than ever."
I agree. I came across a query this morning where the #1 position was strictly an affiliate site, with NO content, just amazon affiliate links, stuffed to the brim with keywords, linking to its self allover the page, keyword match/jammed URL, no domain or page authority, and 0 backlinks.
My real estate site was recently penalized likely do to dup content, we have an IDX feed that adds new pages/home listings daily so it's impossible to get on every page and add my own content to it. What's the best way to handle this issue? Thx
Hi Anthony,
I am working on a real estate site for a new client. We have just finished rebuilding the entire site for him and have installed an IDX feed on many of the pages. The good news is that you should already have the ability to create "SEO-friendly IDX Pages" through your IDX account.
There are several SEO-friendly actions you can perform for your feed pages. These are explained in this IDX Broker Knowledgebase article. They include editing of titles, descriptions etc and sitemap generation for your IDX pages.
If you have any problems working this out you could give your IDX provider a call - ours has been very responsive on a couple of small wrapper issues we needed dealt with.
Hope that helps :)
I've worked on MANY real estate websites, and the best solution is to use the RETS system and build your database interaction from scratch. Too many IDX's seem to be penalized, or are simply unseen by search crawlers because they are houses in iFrames.On two of my most recent websites, https://3bsells.com/site/find_homes being one of them, the custom search feature with RETS has increased their monthly hits from 300 to over 2,000 per month. They also now appear in the top 4 of search results for any address search in Cincinnati.
Does anyone have a WordPress IDX plugin suggestion?
Hey Lucas
Please check the WPL Plugins in WordPress directory. It has a responsive design and support all MLS providers.
It has good ranking, compare the other plugins and I think can be a good choice.
Hope can help.
Hi Cyrus
I have B2B website(trademart.in), can you suggest me how to manage the google ads and content so that not spam by Panda. We have lost so much traffic from april to november.
How we can recover the traffic. Please take a look and help me.
Hey Cyrus,
Epic post, I couldn't imagine a better way of spending my 15min break!
My question is, what about multiple e-commerce sites selling the same product, most of the time the content or product information is pulled from the manufacturer, are they all penalized except for the manufacturer? Any thoughts on this?
Thank you,
Hi TheJake, I work at a largeish ecommerce site and I tested this theory by creating a unique product description on a few items 2 months ago and saw little to no change in ranking. I think google is able to tell the difference between ecommerce vs article sites cranking out 100 versions of the same article. One thing that did help us was implemeting the rel=canonical tag to clean up duplicates between parent and child items.
Hey MarkWrightSEO,
Thank you for your response and I will implement your advice right away.
Though I'm surprised your original product content descriptions didn't improve the ranking for those products. My testing has shown by creating original product descriptions our ranking for those keywords or model numbers has improved significantly. Unless those keywords/model numbers are in an extremely competitive landscape loaded with original content and product reviews. Furthermore, I have this theory that non-orginal content arranged in a orginal format can boost rankings for those keywords/model numbers, your thoughts?
If possible, it is always advised to create unique and accurate product descriptions for all products manually. We have seen over the years that this does make a difference to rankings. As you say, many ecommerce stores will just copy and paste generic text from the suppliers website and when there are 250 instances of this in Google's index, it would be impossible for them all to rank well anyway - which is where 100's of other onsite SEO factors come into play.
The ecommerce sites containing the same set of product descriptions will not be penalised in any way, it's just the rankings (longer tail for the product pages in question) will suffer or not even rank/index at all.
The recommendation from markwrightseo of using the link element will only tell a search engine which page should be treated as the original source and is subsequently only utilised for your own content (it doesn't make logical sense to refer to a 3rd party site shouting "hey, don't pay any attention to my page because that site over there had the same content first".
In some cases, it is not possible to create unique product descriptions, we have one client who has all his ecommerce site/product information served to his domain by another company, and this company does exactly the same for about 50 other sites. In this instance we are having to create a new website as a portfolio site, uniquely optimise and pass highly relevant weight through to his original site and at the same time building trust/authority on his new site to rank that too for a wider range of targets.
There are a wide range of potential duplicate content issues when it comes to ecommerce websites, there is plenty of content online about this. The main causes are query parameters caused by product filters in the categories of your ecommerce website. There are a number of ways to resolve these, one being to block search engines from crawling the filter URL's in your robots.txt - For example: Like the below:-
Disallow: *?price=*
Disallow: *&price=*
Disallow: *?colour=*
Disallow: *&colour=*
Disallow: *?size=*
Disallow: *&size=*
Other ways could include the <META NAME="ROBOTS" CONTENT="NOINDEX, NOFOLLOW"> tag or instructions in your .htaccess file.
Hope that's of some help.
"This is the future of the web – when the equivalent of human eyes look at every page of your site."
What an awesome statement! I vote for that phrase to win "Quote of the Year!". Awesome job.
#5 of the deadly sins mentions affiliate links...
That would mean CJ, Amazon, Ebay, etc...?
If so is there a way to make your link more acceptable?
Is an image that is linked better?
Please explain....
Thanks
Thanks Cyrus Shephard for this engaging google panda article on how to better your on-page search engine optimization to reduce the chances of being penalized by Google.
- Daniel Chege
All these sins are commited by one popular site and didn't affected by any of the Panda updates.
Yahoo Answer.
These 5 sins may seem like no-brainers to the Internet Marketing community, but for the small business owner striking out on his own, he may not realize what he's doing is wrong. There is so much DIY SEO information out there, and not all of it is very good. One SEO expert says black, the other SEO expert says white and the site owner suffers.
Useful post - I am happy that I am not committing any of the aforementioned sins! It does go to show though - most of it is common sense. Write for people and you likely won't go wrong.
Got a bit late here, Panda is all about content, if you can provide high value content to your audience you will never face a problem
"2 – are so dismayed they immediately hit their back button, suddenly popping back onto our radar, alerting us to the fact they were displeased with the result we just showed them – why else would they suddenly come back, without consuming your content? Displeased.
-Duane Forrester"
About this... How do search engines handle when results are opened in tabs. I frequently will open several results in tabs, then browse through them later. Are search engines able to tell the difference? Which site gets "Credit" for the time on site?
great post but it can't solely be the 'template footprint' i have several video websites (ugc driven) but the users rarely add a lot of original content in their description. we encourage them to add transcripts, etc.. but we're lucky to get a sentence out of them. That makes about 70% of our pages 75-90% duplicate, i've done some anecdotal checking and see youtube, vimeo, etc have similar issues although i have no idea what percentage that would be for those sites.
one client had several of the above issues so my guess is that it took an eggregious amount of problems to trigger the panda. they also had an issue which may/may not be unique to them. having used captial letters in their urls (bad idea btw) scrapers grabbed the urls, made them all lowercase and linked to the pages. the problem was the the lowercase and uppercase versions showed the same content without a 301 or other redirect (uh-oh) and for some reason Google never showed these in GWT until we set them to 404 and they poured in by the hundreds.
There are a couple issues with Panda, assuming we're all correct in deciphering it.
- They say too many ads are bad, but you get e-mails from Google telling you you're missing out on "opportunities."
- They say duplicate content is a problem, so how does a blogger post a weekly article about mortgage/foreclosure rates or any other weekly/monthly measure/index of the market without it being too similar?
- They say thin content is a problem, but what about the quick tip, blurb, opinion, message, photo, video, etc.? Or the post that only needs a single paragraph to be the best answer?
I agree with most of the tenets here...quality content deserves to rise to the top, but some of the supposed metrics are troubling.
And it concerns me that we'll all start creating content based on algorithms as opposed to any organic method, which can be both good and bad.
Hopefully more fine-tuning will make Panda a huge win for the Internet.
Its all about finding a balance between these things. Since the spam and adwords teams are completely seperate, they're obviousl going to tell you two different things, b/c they have different agendas. If your sites jammed wiht ads to the point that they are they focus instead of the content, you have a problem.
My take on the duplicate content issue, is that Panda is out to get sites that are scraping content and presenting it as your own, and sites who were tryign to do teh right thing but whos architecture was prone to duplicate content were the unfortunate fall out. Theres a lot of articles on here about keyword research and linkbuilding outreach, but they all find a way to be diffent.
Thin content is only that if its not useful. If your quick tip, or blog post provides relevant information that someone will spend a few mintues reading, maybe even click through to another article, then I dont think it matters how long the article is.
Lastly, I think that Panda is the anti-algorithim update where you can't create content for it. Where past update addressed issues that you coudl find a way around algorithimicly, Panda is a machine learning update based on human responses. The only way to "creat content" for it would be to create content that humans find worth while.
You are absolutely right about Adsense. My site recently went from a PR5 to a 2 and the only thing I can think of is what was said here about ads. What makes me angry is when I started it, I tried to be very conservative with adsense. I had smaller ads, and wouldnt even use image ads because I thought it muddied up the site. But like Truth says, I got email after email from the Adsense team telling me that I was missing opportunities. So eventually I gave in, I definitely saw a huge increase in ad revenue, but now I sit here with decreased page rank which sucks as I used the PR as an opportunity to lure in guest bloggers. I am stumped but I think you answered my question.
Great post, particularly your ending paragraph. Focusing on improving the visitor's experience has to be a priority.
Hi there, good blog and use of examples help.
I am new to SEO and so still learning a lot with a big help coming from you guys.
A couple of points and questions, I would like to make, are these:
For these sites that have been punished by Panda, does anyone know of a site that has managed to get back to their original rankings? and how long did it take them? For me it seems a bit like you have been sent to jail, maybe for something you did that you didn't actually realise was wrong and now you are out, no one wants to give you a job.
For the quote by Duane Forrester, how did the site that was at the top of the rankings get there, if it was that bad that as soon as you landed there you clicked off it?
Interesting correlation however it still concerns me that certain sites have been penalised incorrectly.
I can't wait for Panda to roll out for non-English websites :)
According to Google it did, on (or before) August 12th, except for Chinese, Korean, and Japanese:
https://www.seomoz.org/google-algorithm-change#2011
Just read it @ their blog, thanks! But it hasn't rolled out 100% yet. Matt Cutts recently released a video in which he states it only effects English queries, so it was kind of confusing.
Btw, is it possible to get e-mail notifications from SEOmoz when someone reply's to a comment of mine?
You've got to be careful with Matt's videos, because they film them in batches and then release them over time. Sometimes, when things are moving fast, his statements in the videos can be outdated when they're finally released.
Check the box at the top of the comments and SEOmoz will email you whenever a new comment is added. You will see a snippet of the comment and if it is in reply to another, whose comment it applies to in the email.
You can click the link from the email to come back to the blog and read the entire comment, or just delete the email and only come back when you see one that applies to your own comment.
Nice post Cyrus! The whole point of Panda is to teach people that search engine is a place where people come and find what they want and what they are expecting is the real and quality results instead of spam results that pissed them up in any part of the day…
If you are hit by Panda and need to improve… in general you should change your website to the one that provide best user experience and allow the user to stay and continue their journey on your website instead of bounce back, and to do that once need to consider the points mentioned in the post!
One thing I would like to mention here is that it’s not only about adding a quality content on the website but presentation is also important like how you present you quality content on the webpage.
Over all a great read!
Thanks for another excellent post Cyrus.
I think your comments on sites that are built by machines are actually the key to the inability of many sites to recover from Panda issues. If you approach a site purely as an 'output' or end result of all the things you can do with content elements which by their very nature must be made to conform, the ability to "humanize" that output must become a serious challenge.
The imperfections, non-conformances and unexpected connections that are made by humans will need to be understood before they can be mimicked by automated site builders. Only the most dedicated and determined will look at a mountain like this and choose to keep climbing!
Unfortunately, I fear that some of those faced with this challenge will go looking for easier, less soul satisfying methods of making their way back to the top. So for now, the fact that so many of them are not suddenly back on track is mildly comforting!
Of course, if you have thousands, or hundreds of thousands, of pages to fix and need to apply a little human effort to the process, the tipping point for recovery may be a very long way down the road!
Here in Brazil a lot of people will get in trouble! Nice post!
With regards to step 2 - Empty content - I know in an ideal world you should add great content to these pages to make them valuable but just for agruments sake, if you can't edit those pages what is the best way to deal with them. Stick them in the robots.txt or no index the page?
Great Read, but I've a question. You talk about automated page being damaged by Panda, on our site we have product category pages (which are big data tables) supplying data for each product, so making comparision easier. But these are automated, will these be hit my these updates?
Also does this "ratio" factor mean that our primary content should %-wise out weight fixed items such as menu's and the header of a page interms of <body> code and content?
As you will see we've basically a 2 col layout, but for lots of products we have only a small description (we are trying to improve this constantly but it will never be perfect)... is this an issue?
2 – are so dismayed they immediately hit their back button, suddenly popping back onto our radar, alerting us to the fact they were displeased with the result we just showed them – why else would they suddenly come back, without consuming your content? Displeased. -Duane Forrester
so if I want to sionk my competition I should send them thousands and thousands of crap visitors by buying traffic from some of the IM forums out there? From memory I think you can get something like 10K visitors for about $10 or something rediculous. Sure they'll all bounce, but in this case thats what you want.
Negative SEO for the win? Are we ushering in the age of SEO sabotage here?
I agree with the points you have highlighted but there is so much more to Panda. Google originally set out to deliver the best organic results for any given search and now it's sunk down to whoever can read the minds of the Panda team will win. Since Panda came along it's not unusual to see the one website come up organically on page one several times for the same search. Perhaps the Panda team are getting stuffed brown envelopes from Yahoo & Bing.
We represent several companies with quite technical products, where their prospective clients are focused on quickly determining whether for example polyisocyanurate or phenolic insulation is the best choice for their insulation application. The website must also cater to returning clients who want specifications, MSDS sheets, physical properties tables, and so forth. We attract many visitors due to our previously successful SEO, and we also try to ensure for example that a homeowner quickly reailizes that we do not have suitable products. The websites have been lauded by clients and received national awards. Yet we appear to have been penalized by Panda. We conclude we must now again try to outguess Google, and include video arcade games (insulation-based of course) on the home page to capture visitors even though we full well know will not purchase a product - - while insulting our true prospects and clients by making them indulge the trivia to arrive at the meat. We would greatly appreciate perspective or advice.
Better go jumpt through some hoops and reword all your pages - add a lot of unnecessary fluff with passing relevancy to the page topic so that you aren't panda punched for too much duplicate content on page. Oh, just make sure that each pages "fluff" is placed somewhere that G can see it but doesn't put off regular viewers or distract them from what they really want - and make sure all your fluff is defferent on each page, AND stays within the page title themes so you don't water down your relevancy/consistency too much.
Go ahead and do that for all 1500 pages of your site please.
Don't get me wrong, the overall idea of Panda is good, but the aplication has achieved a "throw the baby out with the bath water" effect in many cases? More relevant results? Only if you like Wikipedia pages.
You have to be happy with a 49% increase in traffic....
Amazing how many sites STILL DO THIS STUFF!
Could you imagine if Google could measure how well people convert and then add that into their algorithm?
I think they should add facebook likes and linkedin shares to their arsenal in determining rankings for websites, and factor in ways to make sure the likes are coming from very active, very unique facebook and linkedin profiles.
And can you imagine how easy this would be to game? buy - return - refund cycles for a #1 in Google! Brilliant!
Awesome post with great examples too! I believe there is still a large dependency with of link-building agencies to keep syndicating slightly amended articles, but Im hoping the Panda changes will keep hitting these. Also good to point out the contrasting noises coming from Google on Ads.
Great post. Out of all of these, reducing your "footprint" is by far the hardest. It's especially difficult to have to tell a client that they're going to need to scale back on their advertising, particularly after they've lost a ton of organic traffic from Panda and are bleeding money.
How about User and Machinery Generated Content like scribd.com and slideshare.net ?
their content are mostly machinery generated (rendering document file into text). They seems have many overlapping content and keyword on their site. But they both add value to user.
Do you think panda hate them?
Let me know how you think about this. Thanks in advance.
Very insightful... if I may summarize what the Panda update is, it is all about great user experience. Great user experience means exact relevance of the user's search keywords with what is displayed on the resulting SERPs. That means -- good relevant, un-rehashed, quality content.
On the google business side -- satisfied searchers, translate to satisfied customers, resulting to great business.
That's why Google is a monster billion dollar company!
Stop the sychophantism my friend. Not to be rude, but while the theory of your idea may be true, the practice is anything but. The internet is littered with great, relevant, unique content, much of it by personal experts blogging on their personal sites.
Problem is you never see any of it because page 1 of the term you may be seaching for has a 12 pack of sitelinks from 1 site. either that or Wikipedia, about.com or some other heavily internally SEO'd site has a half arsed article on the topic but ranks better because of the domain it's attached to.
G has no real, reliable way of evaluating the "quality" of information on a page other than spelling/grammar perhaps. This is why a crap article on a strong domain will still show up in #1 (doesn''t matter if it's accurate or not, as long as it's unique more or less) over an informative, accurate article written by an expert on his personal blog if he has little SEO training.
worse yet - cosider this: You infomative well researched article is posted on your site, which is regularly scanned for content specifically because you know your stuff, and reposted on a larger, more prolific site word for word. G will probably index their copy first as content can turn up in G's index literally minutes after posting, then label your copy (the original) as a duplicate when they finally find it.
Nice Info, but you spoke about duplicate content on our pages, then sent us to a site that checks duplicate content between different sites. Do you or anyone know how (hopefully for free) we can check for dup content withing our own site?
I believe SEOMoz provides this as a part of Pro account tools. You can add your websites and get automatic reports on duplicate content within your website. Google also added this recently under diagnostics in Webmaster Tools.
With much respect, I see no evidence here that SEOmoz was impacted at all by Panda. I have a customer that lost about 50% of their traffic after the first Panda algorithm change. If you're going to write a post on "Beating Panda", you need to write it about a site that was actually impacted by Panda. SEOmoz benefitted (as did others) as Panda-penalized farms were dropped from the index. That's not because of what you did or did not do... it's because of what those sites didn't do.
I'll throw you a bone:
Oh and in regards Ads: What do you think removing ads will do to any search engine bounce rate metric (let alone your ad revenue?) Tread carefully here and don't necessarily believe the hype.
Good Luck :)
If bounce rate becomes a serious metric we're all in serious trouble. Negative SEO (factors that can HARM your serps rather than just not help) is a slippery road to an ugly, ugly internet. How long do you think it'll take for a bunch of spammers to set up a "bounce network" which essentially esists to up the bounce rate of competitors sites and thus harm their SERP's?
Would you like your prize site, which you've poured year into, suddenly smacked back to page 15 because a bunch of a-holes decided it was easier to know you out of the serps completely than it was to out rank you?
I'll let you do the thinking about what you'd do then? Reinclusion request? See you in 6 months!
I like the idea of having no ads on a webpage - if you need them to run said site, might want to rethink the business model.
Great article, new update to it coming soon?
Great post Cyrus! You offered link to duplicatecontent.net. I played with it for a bit and browsed the FAQ. At that site, there are several metrics given in two groups: HTML similarity and text similarity.
I ask Mozzers:
For example, is a total text similarity of 85% or less between two product pages "good enough?" 90% OK? Some other metric/value more important?
What's the advice you'd give to clients using the tool to check that they have enough original content on each page?
I look forward to reading responses. Thanks!
I have a very large website that has lots of unique well written content and not copied from other sources.
We also have a comparison shop, which pulls in prices, images and descriptions from retailers/merchants, which of course is duplicate content, but in my opinion provides the website visitor with a very useful services, allowing them to find prices easily and save money, they can also navigate easily between the comparison shop to other website pages, that as above are unique and well written.
I would love to hear your thoughts on the above, based on the Panda, which can penalise sites like ours, regardless of whether it is providing something very useful. Why should we be penalised for providing something useful to users?
The comparison shop is not a general comparison shop for everything and anything but a target niche, although a massive niche
Thanks
Thank you for the post Cyrus. Is the most focused post about the panda update I found. Everyone should read it!
Nice post Cyrus Shepard, I want to tell you that I am self taught SEO and have been learning through video tuts, blog posts and forum discussions online. Its been over 3 years but after reading your post I learned some very deep points and I want to thank you for this interesting post.
Hey Cyrus! The nifty "duplicate content tool" you mentioned in point #1 is hyperlinked to a very misleading site. Was this a mistake, or am I just missing something?
That was actually pretty awesome! i love this article
I personally am not convinced about the longevity of the panda update.
There is little doubt that in the space of less than a year the quality of the SERPs has imporved drastically, as spam has been hit with a real curve ball. But, as so accurately pointed out by so many, it stands in opposition with the google adwords team. The adwords and spam team are in totally seperate departments, and it is likely that Matt Cutts' team did not involve the adwords team as much as they perhaps should have done. How else could the rug have been pulled so rapidly from under the feet of the adword team?
Fact is fact: websites too crowded out with adwords are penalised by Panda. But remember that Google is an ad-funded enterprise. Over-advertising websites is big business for them and it is not financially viable that they attempt to sideline the very websites spending the most on advertising.
I feel that Matt Cutts may well be told that he has, once again, done his job far too well. He has dealt spam a serious blow, allowing Google to do what they are there to do: produce the most relevent results. But returning these results is only part of the picture. They also need to make money!
Perhaps in the coming updates we will see a compromise on this, with websites with heavy advertising regaining some of their lost ground. Either that or there will have to be a major shake up in the adwords department, as they look to find ways to make their advertising more presentable and less of an eye sore.
Watch this space methinks...
Google Panda 2.5, Give more importance for fresh contents
Thank you. Laets Panda Update affects many coupon code and shopping sites, especially Amazon. It means overlapping and related content need to be update or must removed. Video sites getting value than article news sites.
Thanks
[link removed]
Ditto - a really helpful tool - although equally great to have the duplicate content flagged in the Pro Campaigns too.
I love stats in this instance - does Google now think you're providing a higher quality of content over the course of the Panda updates or are you supplying more content of the same high standard - thus increasing your chances of gaining more organic traffic?
Either way, glad Google agrees that your content is great and we look forward to having more of it :-) Have a great weekend.
Panda scares me a little bit - because with all the efforts you make when creating your sites you never know what's considered to be grey hat. This article assures me, because I don't do most of that stuff that's flagged...
Thanks for the content :-)
Interesting post. I am glad to see somebody address the discrepancy between adsense recommendations and search rankings. I have suspected this for some time, that their right hand and left hand are not on speaking terms.
I went to an adsense in your city conference about a week before panda hit. I was struck by the youth of all the google employees, all 20-somethings with most of the publishers 40- and 50- and 60-somethings like myself. I think that google teams do a lot of experimenting.
Personally I am still confused by the results I now see on google, touted as "better" but certainly not better for me. When I make a search, I want exact keyword match, not popularity, and certainly not several results from each of the top sites, usually led by wikipedia. It would be nice if there were an option for that, what? If you could select what type of search you want?
I actually use bing sometimes now...
Oh well, color me naive...
Great article. In essence, most of the points mentioned in the article is common sense. But as the saying goes, "common sense" is not so common ;
The lesson to be learned is to not put all our eggs on one basket and diversify our traffic sources. You don't want your business to be solely dependent on the mercy of Google gods.
It's time to do guest blogging, ezine advertising, blog commenting, video marketing, podcasting and forum marketing so that traffic comes from a wide variety of sources.
Very nice sum up...Panda effects mostly on ad placement ratio,affiliate link placement ratio, thin content and dupe content page's...
Though since several version of Panda running..still I do believe there is no solution that a pandalized site can FULLY recover...seems Google is trying to fill up serp with brand and new source of information...
Great post with some quick tips. A lot of the "how to beat Panda" posts I see are tl;dr - this is a great, quick read that's easy to share with others. Thanks for writing.
Good tips and nice graph on progress. It's nice to see good sites with good content getting more traffic. This means they are on the right track and it will take a little more time and data before they get it all right. Thanks for the tips and tools and keep up the good work.
There is a big gap in this article: The impact of spam attacks on a site that allows user-contributions (e.g. discussion threads, wikis, etc). From my experience, not taking down a spammer's attack will have drastic effects on your rankings. It poisons the entire domain. But, the logic of how fast to take it down, how much is acceptable, etc, is very unclear. The correlation between an attack and a drop in traffic from Google is extremely high. And, recovery takes a long-long time. More needs to be researched on that topic. Any other sites getting the same scenario?
"1 – are so engaged they forget about the rest of the Internet for a few minutes and bask in your glory, getting the information they wanted, and more."
Sadly, in my case... this is proven 100% dead wrong. we have 20 minutes average on page come up on page 3 consistently while the reports I have seen of most page 1 competition is less than 5 minutes.
Nice post as always, Cyrus. It's good to see that the SEO Moz traffic is doing so well; I'd be curious to know whether the uptick in traffic has been based on long-tail or head terms.
When I read a quote like that from Bing, or a suggestion that that's how Google judges quality, it really makes me think that most people browse the web differently than I do. Do you open ONE window at a time and then hit back? Seems inefficient. I'll open anything in the top 5-10 that seems to match what I'm looking for in new tabs, and close most of them immediately. I'm probably ruining Panda. :)
Most of our traffic is long tail, and we've seen a proportional increase in traffic along that front.
Hi Cyrus, good point, I do have a question regarding the duplicate content tool you linked in the post. The website itself doesn't provide much info on how to read and interpret the values shown, I was amazed at some of the results and they didn't make sense, so I compared the link building category from my blog with the SEOmoz link building category, and the results I get are: 55% total html similarity and 40% standard text similarity vs 30% total text similarity.
The smart numbers are the only ones that make some sense, 20%... so any input on this?
thanks
Very useful. Thank u
I like this post, except the first graph. As for me Panda update influence for branded search is about 0. And according Goolge insights for Search Interest to "seomoz" increased about 50%. So nothing strange in 49% total search traffic growth. https://www.google.com/insights/search/#date=today+12-m&q=seomoz Non-branded search will be more informative for Panda's reseach)
Smart, very smart question! Non-branded traffic was up over 50%. Non-branded search also makes up about 80% of our organic traffic.
So how long until Google starts ascribing value to your site based on bounce rate, and time on site? Are they doing this already? This seems to be a very strong indicator of how valuable people see your site.
Good post, thumbs up, etc.
Hi,
i have a question about Rule No. 2 "Empty Content".
What you guys exactly mean with "Eliminating these types of empty content pages"?
Is it just fine to us "noindex, follow" to get those pages out of the index or is it nessary to delete them.
Because, if I use "only" noindex,follow the pages is still "there" and crawlable - even its no longer in the index.So, could it be that this site will still be Factor for the Algo - even its not in the index?Please excuse my bad english.I hope you understand my questions.
Thanks!
I am presuming that you have a 404 coming up? If so and even if not, 301 redirect the pages to the homepage.
I'm not sure how similar Dr Pete's example image is to your site, but if it's similar then add something useful to the page - there's no point having a list of page numbers on a page. Why not add a list of 'Most Popular' pages underneath, a call to action, a fast loading image...anything to prevent the visitor feeling lost or thinking that they've hit an error/the page isn't loading properly.
Thanks for sharing that duplicate content tool. It is nifty.
I think the graph at the top is very important. It shows that users who have regurgitated SEOMOZ's content else where have been stopped by the panda and the traffic is now being brought back to the seed of the content where it belongs.
Its kind of inspiring to write better content as I now know, I will benefit from my content rather then someone else.
"Google even tells us to plaster ads all over our site" - to be fair to Google, not in the link you've provided they don't - and I haven't seen them do that anywhere else. The image you use is a heat map suggesting where the best places might be. They say "While this heat map is useful as a positioning guideline, we strongly recommend putting your users first when deciding on ad location."
Otherwise, a useful post.
What should a large meta search engine do? If you see, all links on a listing page end up leading to affiliate partners?
The intent of a user landing on that page is seeing listings and any website cannot ignore that very important factor.
Suggestions?
Thanks for sharing these tips. It's our belief that user experience and engagment are the foundation of a website's design, content, nagivation, etc. Panda, and your advice, reiterate that belief. Again, thank you!
If you're churning out poor content purely to generate money from ads and not actually provide a service people want you get what you deserve IMO.
When it comes to ad placement - there's a good reason Google Adsense only allows you to place 3 ad blocks on a page - if you place too many, you dilute the content on the page and your ads will perform worse overall. Don't just surround your content with ads - create value and place them smartly.
@cyrus - good post. Is there any correlation between your traffic increase and a decrease in your competitors traffic?
"Adwords says YES, Panda says PENALTY" - it's a very clever system really. Without SEO people would have to resort to Adwords for the majority of their traffic!
Great post and some really good examples too. There is still a large dependency with a lot of link-building agencies to keep syndicating slightly amended articles, but hopefully the Panda changes will keep hitting these. Also good to point out the contrasting noises coming from Google on Ads.
Yes I agree with the point about having layouts which are non spammy, it is really key to be driving layouts which are high on unique original content to keep the users happy and also the search engines happy =)
The sites I have seen personally which do not fare to well after panda are sites which are aggregators of content.
I agree with your post. Great post! I think today, make websites cleaner and content rich so that it would be appealing not only to the users but to the SEs too!
Panda is just another change to enforce what we should have already had. We all knew the value of unique content, no empty pages, etc. so it shouldn't come to a surprise to the people who got penalties.
I will be seen how things go since Iam new to this.
I feel for those that are getting panda slapped, but I for one am all for humanizing the internet. As google gets smarter I think the pages we land on will be more useful and entertaining as a result.
It also makes the jobs of SEOs easier. Optimize for human eyes! (heh I like the sound of that, hmmm...)
If we don't have to split our focus so much between developing content for search engines and developing content for users then everything is simpler.
Great blog. Here's an interesting thought re: Duane Forrester's Bing quote:
In the UK sites now have to explicitly get permission from a website to place cookies on their browser. It is expected that the next year will see browsers moving towards having 'opted out' as the default option, with users then having to amend their settings to enable tracking.
I am assuming Bing uses cookies to track the behaviour of users in Duane's example. If that behaviour is used as a strong metric, I'd be deeply concerned about the impact of the recent cookie legislation on SERP quality and subsequent site performance.
Good post. I think 'duplicate content' is a big sin too. Many sites knowingly or unknowingly duplicate their contents and if large number of pages on the site are duplicate then the website can be considered as a low value website by google & get the panda slam.
I'd like to believe that since so many of these sins are caused by templates and content management systems, they can be solved by smarter templates and more thoughtful CMS's. There is actually good content in (some of) these sites. The solution may be found in organization and human-based presentation.
@Cyrus: what's your definition of smarter templats and more thoughtful CMS's? Given that the actual copy, images, videos are original, If some is running a regular template on a mainsteam CMS like Jommla or Wordpress, what could be done to minimize the risk of being labeled as duplicate site?
yes,cyrus you are right..the tough thing is to beat panda from a good templates and better Content management system.And anohter main thing of panda update is content farming and thats the reason,many sites have been flushed from SERP...
seo-himanshu and all...quick question about duplicate content:
On Wordpress/blog pages it kinda seems like the tag pages cause some inadvertant duplicate content. I think my Webmaster Tools shows the tag pages as independent pages sometimes, and of course they will have duplicate content simply because they are combinations of blog posts.
Is this is a problem, or am I being a noob?
Thanks!
I'm going to go out on a limb here and say I don't believe CMSs and templates are major culprits. If your template dominates your site, then yes; if your pages are thin on content, then yes.
In a CMS/template environment a certain percentage of your code is always going to look similar (in the case of default WP templates your header, footer and sidebar are already the same, and that doesn't even include all the other code that goes along), but if you're only going to sit back and do nothing about it, you deserve to get nailed.
I did a little test using the dupe content tool you linked to - I run a WP blog - my current setup has article snippets (first paragraph or two) on the front page, not full articles. So I checked my front page vs my latest article (example.com vs example.com/latest-post).
HTML similarity - as I expected, was reasonably high (around the 71% mark), but text similarity was at about 47%. Personally I feel the text similarity would carry more weight as penalising factor. HTML similarities are somewhat unavoidable.
In the end it basically comes down to the actual useful content being important.
@BartlettPair - I don't know if anyone will actually agree with me on this - but I generally avoid the use of tags on my blogs - mostly because of the possible dupes. If a search function is really necessary, I embed a Google CSE rather than use the WP variant.
Tag pages have a lot of user benefit so it's not worth removing them from the nav altogether imho. You are right on the duplicate content issue though (who would want a tag page to rank?) - the best solution is to robots "noindex" those tag pages so they're still there but won't hold any ranking ability.
However I would say definately leave off the "nofollow". That way the pages can still pass *on* the juice they receive to the important pages.
Best of both worlds - Jon
I would agree with you on this one.
I have observed significant variation of traffic on site that have duplicated content on different ccTLD.
I found that Google tends to be very confused about the whole thing and from now on, would only recommend diversification strategy for SEO players with big pockets already dominating their niche.
Working on a post about this. Stay tuned.
@Cyrus,
Yes definitely proper content handling via advanced CMS sorting although I would still be cautious as most companies would use shortcuts to avoid re-investing in localised content.
Sorting a review, blog post, product or text snippet would still be considered the same by Google IMO.
I believe it is best to consider the localisation exercise as a whole new site. Makes a lot easier than you think.
re: duplicate content on different domains. I agree - we run 3 high traffic sites - all with the same content. All have a different look and feel. All are highly dependent on Google for audience. Here is what we found:
-the site that has few backlinks got hammered, losing 60% of our traffic when the new algo hit. Interestingly, this site should be seen as the canonical source of the content. We publish there first, we have some cross-domain canonical tags and attribution links pointing back to this site from other sites, etc..
-the two sites that had high quality and numerous backlinks didn't get touched AND have recouped some of our losses from our other site that got hit hard. On many articles, these sites point back to the site that did get hit. Hmmm.
I would argue that, visually, the site that got hit was the best site - though maybe I'm missing something. For us, it seemed to be the traditional factors that got amplified with Panda, i.e. backlinks.
I dont think it has anything to do with duplicate content cos if that was the case the entire e-commerce type of sites would be gone from google and that also includes amazon and ebay. All the content comes from manufacturers and everyone use the same stuff.
Yes, it may be that google have been cheeky in what they are doing. Since they can't afford to hurt players like amazon and ebay they might have "overlooked" them !!!
At the end of the day its all about the bottom line. Google is not into charity, that is for sure.
Great article. Obviouslly there are sites that seem to "contradict" Panda's update (mostly large brands). But Google is trying to give real people, the results that they can benifit from the most I think they are doing a good job, and I think this is a good thing.