22nd February and 11th April, 2011 – I will remember these two dates forever. One year ago, Google pushed out a much needed algorithmic update to its search filter, popularly known as the Panda farmer update. This algorithmic update, as you might know, was designed to target content farms and spam sites but the real scenario was very different from theoretical goals as thousands of legitimate publishers saw a big dive in their organic traffic. Panda was launched to address “low quality” or “thin” content and being a site wide penalty, this update did not make any exceptions; changing SEO best practices forever.
The reality: Panda is color blind and sees things only in white or black. Either your site falls entirely in the Panda pit or you are never affected at all. There is no hanging in the middle, pure binary stuff.
The most frustrating thing regarding Panda is that it is a rolling update and the iterations are manually pushed. You can’t expect results overnight. You have to patiently make a change, wait for 90 days, see if your changes were noticed by Google and move on to make the next change. If your site was affected by any of Google’s Panda iterations, you know how difficult it is to arrive at conclusions. There are a lot of factors, some examples:
- Site design, architecture, navigation and usability.
- Advertisements – how much is too much?
- Content quality – the definition of “Good content”, “Shallow content”, “thin content” and “poor content”.
- Content exclusivity – why does your website even exist?
- Links / Social signals.
The Panda Timeline
Here is a brief timeline of the entire Google Panda update saga:
- Panda 1.0 (February 22, 2011): Google Officially declares the launch of Panda algorithmic update.
- Panda 2.0 (April 11th, 2011): Google Panda goes global for all English language users, affecting 12% of all queries.
- Panda 2.1 (May 10th, 2011): More iterations of the ongoing Panda update are reported.
- Panda 2.2 (June 16, 2011): Panda continues the devastation
- Panda 2.3 (July 26, 2011): Google acknowledged that a small update was pushed to the Panda filter. Big shakeups reported, traffic to a majority of sites disappeared in thin air.
- Panda 2.4 (August 12, 2011): Panda rolled out for major International languages excluding Japanese, Korean and Chinese. This was the biggest impact after Panda 1.0 and Panda 2.0, impacting 6-9% of all queries.
- Panda 2.5 (September 30, 2011): Another iteration, this time confirmed after Webpronews broke a story on Daniweb.
- Panda 2.5.1 (October 9-13th,2011): Finally some hope, Searchmetrics reports Panda 2.5.1 resulted in major recoveries. Google Engineer Matt Cutts confirms.
- Panda 3.0 (October 20, 2011): A minor update.
- Panda 3.1 (November 21,2011): Another small update affecting less than 1% of all queries.
- Panda 3.2 (January 18,2012): Another minor iteration.
- Panda 3.3: The recent Panda iteration changes how link popularity influences search results.
Only a few websites were able to recover from Google’s Panda slap. Digital Inspiration, Daniweb, CultOf Mac, Pocketlink are some examples though none of these sites were able to completely recover their lost organic traffic.
A Case Study And An Honest Analysis
First things first, I am not a search engine optimizer and neither I specialize in Panda recovery techniques. I am not a pinch of those super Probloggers out there, I can’t help your site escape the Panda squash. And I doubt anyone else other than you can.
But I will share a case study of my site which got hit by the Panda 2.0 update and made a partial recovery 2 months back. I will share the things I did and the things I did not do, in order to recover from Panda’s grip. Remember that every website is different and what worked wonders for me, might just be disastrous for your site. Never generalize!
When Panda stuck my site, I instantly lost 70% of the traffic in one shot (by traffic, I mean organic traffic from Google.com). Following subsequent updates and iterations, the organic traffic was highly shaky (40-60%) and I have had nightmares thinking about how to fix this problem. Needless to say, Panda drove my site’s revenue further downhill. Here is a screenshot:
Here are the things I did, without having any success at all:
- I read this story at Digital Inspiration and found myself in the same boat. I deleted all the useless tags, categories and kept only the ones that are essential for site navigation.
- I deactivated my site’s mobile theme, thinking that it might be giving rise to duplicate content, cloaking and so forth. This doesn’t make any sense now but when your ship is sinking, you have to try every god damned thing!
- I started filing DMCA’s in bulk, taking down content scrapers, auto posters and spam sites. I used data from Google Webmaster tools and nuked each and every site that copied content from my blog. This was an enormous job but I did not have any other option.
- I went all guns blazing towards directory sites, killing every duplicate trace of my site’s content from the web.
- When my original content was outranked by scrapers, I made fun of Google search. Time and again. And again. Until I realized that blaming Google won’t help at all, I have to drink my anger and do something that makes sense.
- After a few weeks, I realized that I am on a wild goose chase, going nowhere. I hired someone to file DMCA’s on be half of me while devoting time in fixing other site specific issues.
- No-indexed all the tags, category, archive, author and search result pages. Any page on the site which was not required was deleted. Any feature not exclusive, was forcefully killed.
- Cleaned up the FTP directory, got rid of junk files and disallowed bots from indexing unnecessary directories of my site.
- I discoveredthat a lot of my blog posts were shared on Facebook as “Notes”. A majority of these notes were auto –created by myself and in all cases, the visibility of these “Notes” were public. While performing dummy searches in Google, I found that these notes were showing up on search results and my original blog post was no where to be found.Guess what? I killed all the Facebook notes within a week. When I think of this action now, I can’t stop laughing but like I said, “when your ship is sinking, you will throw your arms in every possible direction”.
Until November, nothing significant happened. Traffic had stagnated, nothing was working. I talked with fellow blogger friends and found that none of them have achieved any ground breaking solution to this problem.
Then I thought, maybe I should review my old content and see if it needs more attention and human care. Here is what I found:
- A lot of my old blog posts were either outdated or made no sense at this point of time.
- They were very short in length (150-250 words if not more).
- These posts were not “Exclusive”. Simply put, they were nothing but “rehashed content”, which I wrote after getting inspired from other sites. (Every other blogger begins his journey like this, reading through his RSS reader and putting the same information on his site)
- These pages were written very badly and were full of grammatical mistakes. It was evident that the writer was in a “state of hurry” and wanted to hit the “Publish” button as quickly as possible.
I have no shame in admitting this, as this is the truth I must face. When I started blogging from ZERO, my writing was outright horrible. Honestly, it is still not very polished and neither cent percent accurate. But comparing to those posts which I wrote back in 2009, it is at least readable and makes sense to some extent, no?
A big shout out goes to Mr. Clif Sipe, Udit And Amrita, who painstakingly read each and every post I write here at Techie Buzz. Without these guys, I could never have improved my writing in the first place. Heck, I never knew the differences between [it is, it’s and its]; every email I receive from them is full of red marks all over the place. I am sure this epic rant of mine will have more than a thousand mistakes, so whoever in charge of editing this post, please spare my nonsense. One last time :-)
After discovering tons of badly written pages on my site, I decided to check how much organic traffic these pages have received before the Panda algorithmic update was rolled out. Here is what I found:
Above one is an example page which got only 8 page views from Google.com for a period of one year before Panda was rolled out ( Google Analytics > Site Content > Pages > Filter by URL). Notice that although the page has strong user satisfaction signals (low bounce rate, high time on page, low exit ratio), it was not performing at all.
Needless to say, the page was poorly written, without thinking whether the user would read after the second sentence or not. The blog post (now taken down) also had a few broken images and the provided solution no longer worked. This is a “thin” page and my site had these kind of pages in plenty
Enter – The Supplemental Hell of Google Search. A supplemental page will still rank in search results, but only if there are not enough pages in the main index that are returned within the search.Google used to place a “Supplemental Result” label at the bottom of a search result to indicate that it is in the supplemental index; however in July 2007 they discontinued this practice and it is no longer possible to tell whether a result is in the supplemental index or the main one.
I thought – “Okay, maybe I should get rid of all these pages and see how things change from here. If they don’t, I will fall back.”
I remember the words of Google employee Wysz:
Bear in mind that people searching on Google typically don’t want to see shallow or poorly written content, content that’s copied from other websites, or information that are just not that useful. In addition, it’s important for webmasters to know that low quality content on part of a site can impact a site’s ranking as a whole.
Here is what I did next:
- Installed the Broken link checker WordPress plugin, fixed all the broken links and missing images.
- Used XENU’s linksleuth to get rid of all missing images from a post. The result was that all the broken images were either replaced by new images or removed.
- Signed up at Grammarly.com (premium) and found out all the pages which were full of grammatical errors.
- Redesigned the site from scratch, moved content above the fold, got rid of Kontera, Infolinks and other in text nonsense.
- Installed the TD Word count WordPress plugin – this little script lets you arrange all the content of your site according to Word Count.
- Backed up the My SQL database, restored the current condition of the site under a deeper directory (blocked via Robots.txt). Now I am ready to do a full house clean up.
In the process of cleaning up, I deleted close to 300-400 posts. Any page which was short in length, written badly and never performed in Google search before Panda, was killed. No redirection, I left the broken URL’s 404, as advised by Google engineers. If I found that a page makes sense and there is scope for improvement, I edited that page and added more insights wherever possible.
The result: A slow recovery but hey, I did it. My traffic returned, not entirely though.
You Will Be Affected By Panda If ……
1. You churn out (or used to write) 6-10 blog posts every single day, averaging 150-200-300 words.
2. You write about software reviews blindly by reading reviews posted on other tech blogs. Developing content, merely by reading someone else’s review is nothing but madness.
Example: Some company released a tiny little script or app, you got an update in your RSS reader and thought it should be blogged right away. You download that app, use it, write a review and move on to writing the next post. Typical short review no longer than 250 words, a few screenshots and download link at the bottom.
There is nothing wrong in short reviews but at the end of the day, are you providing anything additional? There should be a marked difference between your blog post and the blog post of the app developer. If you can’t add any additional insights (other than linking to your old pages through out your post, which I think is another form of spam), why write this review? What purpose it solves?
If your site has tons of original content, lengthy pages with detailed text, these short reviews wont hurt. But if your site has tons of short reviews and only short reviews, you might get hurt real bad (depends from one site to another site).
3. You pay no attention to grammar, never proofread and believe in the volume of content.
4. Needless to say, you will be affected by Panda if your site has duplicate content or copied text from other domains.
Exclusive Content Goes Way Beyond Your RSS Reading List
Your site must have some exclusive content. I said some because practically, it is very much impossible to produce exclusive stuff every single day. But your site should have something exclusive, examples include detailed reviews, editorials, opinion posts, research and resource pages. If you are dumping short reviews and only short reviews in bulk, you’re adding to that echo chamber of spam, please don’t do it.
I know, I know. I know exactly what you’re thinking. That majority of posts at Techie Buzz are not exclusive and they are already available on other tech news sites. I will explain that.
Here at Techie Buzz, we write a lot about Tech News and breaking stories. But we just don’t write or rehash from other news portals, we add our own insights, commentary and opinions to the topic. News has to come from somewhere and majority of Tech News sites work the same way as we do. We can’t produce news from our embryo, as nobody else can. Our posts are regularly featured on Techmeme (example), I can flood this post with tons of links and citations but I would refrain from doing it here. “Blowing your own horn” is beyond our ethics.
My whole point is simple.
Don’t re write the same thing over and over again. If you can’t add your own insights, commentary or opinion about a topic, don’t write it. One original thought is much better than tons of garbage – the sooner you realize this, the better off you will be.
Finally, the panda video.