Google Panda Recovery – The Journey So Far

Google Add comments

Welcome to part two of my series of an unknown length looking at my exciting adventures recovering one of my affiliate sites from it’s Google Panda woes!

The first thing I did after returning from New Zealand and getting over the huff that I’d been in was formulate a recovery plan.

Here’s The Plan Stan!!

  • Removing all site content except the core pages around which the site is themed and rebuilding the whole thing from the ground up. In a site of almost 1,300 pages this leaves me with around 150 to play with.
  • Increase average page views per user, time spent on site, and general visitor happiness by not having any core pages that redirect straight out to merchant.
  • The core pages (which feature individual fashion brands) will be completely redesigned to form decent information resources about the brand beyond the current description and “shop now” type links. Things like sizing guides, information on the types of product they specialise in etc. I’m also going to have a special offers widget created for the site that’ll allow me to present information on sales, discount codes, and offers for each brand or by specific category.
  • Investing heavily in site functionality. I have an early version of a plugin that’ll let me take merchant feeds, grab and save all their images, resizing and renaming them automagically (Oh how I’m coming to love that word). The plugin also creates masked affiliate links and will allow users to browse all the products in ways they wouldn’t be able to on a single merchant site. Each core page will have a small widget which will display the latest products on the site for the brand. I started organising this back in January when I realised that we had a problem.
  • Thin “product reviews”  which were really only rewritten product overviews and whizzed people straight back out to merchant will be replaced by newsy items and how to articles.  These will not be monetised.
  • I’ll also start doing real product reviews under some kind of nom de plume.  We’ll be setting up a mini photo studio to take detailed pictures of bras. I will not be modelling these :D. These  reviews will also form part of my shiny new social media strategy. I shall be trying to get the merchants and brands to bestow some retweetage upon me for my efforts. Again, these will not be monetised but I’m holding out real hopes of scoring serious amounts of free underwear and link love once I get it going.
  • Competitions – again around social media to try to get people retweeting, liking, and generally engaging with the site.
  • I’ll be going through and re-optimising the core pages, well de-optimising them really. I’ll also be rewriting anything that isn’t clear and concise and where there’s too much repetition of terms closely related to the page content.
  • Lots of social media type fun, all totally new and nosebleed inducing but I think essential for the longer term.
  • Theme update.
  • Site speed overhaul.

If you read the above carefully you’ll see it caters for an awful lot of things Google could use to tell you are an unoriginal site.

It Is Alive!! The Early Results

I’ve already put in two solid weeks of work on pulling my site apart and starting to reconstruct it and am seeing some early, but limited, results.  Here’s a list of what changes I’ve made, when I made them, followed by some results and observations.  Hope you’re still awake by the time you’ve gotten through ’em!

First of all, here’s the Google referral data from my Analytics account: –

Encouraging hey?

The upward trend is all the more pleasing when you consider that the increase in traffic is being countered by the removal of all the pages that were generating 90% of the traffic we’d regained as at April 25th or so. In a twist of fate only Google could engineer the only pages left ranking were complete and utter spammy shite which I’d accidentally let get indexed – they were old PPC pages and had no unique content whatsoever.

What Have I Done So Far From My Plan?

I hear you ask. Here’s a list of what we did, and when.

  • April 3rd. I changed all the brand overview pages so that rather than the user going straight out to merchant from my “view whole range” button, they were directed to my internal product feed section which I thought would increase the average page views per user that Google was seeing happen on the site.
  • May 2nd. I removed all the content from the site except my main pages. This took my 1,300 page site down to a whopping 169.
  • May 4th. New site design and structure is put in place. For now, I’ve based the site completely around the core content. Replacing the old feed content in the top bar with links to categories containing all my brands.
  • May 6th. I began the process of deoptimising my pages and rewriting content where needed.
  • May 12th Finish the deoptimisation and rewriting. I can assure you alcohol will be taken this evening!


  • Like Site B I found some pages on Site A where my SEO plugin had let me down. A significant number (but not all) of them were quietly ranking away. Not enough to alert me by producing much traffic, but high enough to cheer me up significantly.  This reinforced the idea that softening the SEO on the site was looking like a winner.
  • As I went through my pages on I also did test searches and discovered that many pages that were previously totally buried or around page 5 were now mid page 2.  Another trend was that in many cases “softer” site pages that perhaps mentioned the brand name once were ranking relatively high in the serps i.e. within pages 2 to 5.
  • The cached date on the majority of the pages returning to the rankings was initially around April 14th to 18th.  So either my change to the way people flowed through the site early April was something Google liked, which has resulted in some pages at least appearing within the first five pages (they weren’t before!) or I’m simply benefitting from an algo adjustment and this information is meaningless.
  • Many of the pages I adjusted on were reindexed immediately. Some of them were suddenly looking a lot sunnier on the rankings front. Others still nowhere in sight. An oddity with this is that the new titles and meta descriptions are showing in the index but the cached version of the page is still old.  Another thing I’ve seen is that pages I’ve adjusted will often be catapulted back into the SERPS but it’ll be the old page version ranking rather than the new one. I’ve seen this repeatedly in the last two weeks.
  • Average time spent on site and visitor page views have more than doubled and are now in line with my other sites.

Wow. Well this post has got a bit on the long side but it’ll hopefully be useful in this form.

Before You Get Jealous Of My Recovery…

Green eyed SERPS monster bothering ya? Never fear!

Because I still have a very, very long way to go.

With having made the decision to take the site back to it’s core I’ve lost a large number of pages that previously brought traffic and income to the site. I decided that for my site there was no quick solution because when I looked at it and asked “does this add value and could it stand alone without the affiliate content?” The answer was a resounding no.

I think it’ll probably take me at least 6 months to get back to where I was but I honestly believe it’s more than worth the effort. I know I’ll end up with a much stronger resource that really does add value in the long term – which will put me streets ahead of those not prepared to put in the work.

Good luck to everyone out there who has been Pandalised – and remember the road to recovery isn’t an easy one but it is worthwhile!

Related Affiliate Marketing Posts

27 Responses to “Google Panda Recovery – The Journey So Far”

  1. Jon Cook Says:

    Certainly looks like you’re going the right way of tackling this. Building a site up over so many years you can quickly forget about the crap that started out on the site and what you’ve learned. By starting afresh you can take stock of what content you currently have, what works and what Google currently sees as good practice. This is really difficult to do with a site that is ranking well already as you’re scared to slip down the SERPS however after Panda I guess you felt you had nothing to lose?!

  2. Kirsty Says:

    Hey Jon – yep I had nothing to lose. The site was getting that much traffic it was too scary to start messing with and I was really unhappy with it as it stood.

    Hopefully my detonate and renovate technique will pay off – I don’t expect a quick fix with this, it’s a long term strategy!

  3. Jon Cook Says:

    At least your other sites will help keep your income levels up Kirsty, that can never be a bad thing!

  4. Kirsty Says:

    I’m very lucky in this Jon, I’m still making the same money as I did in the same month last year. I worked very hard to make sure I didn’t have all my eggs in one basket – and thanks to two other sites that have risen fast over the last year I’m in a position to do this properly without panicking!

  5. Gary Says:

    Hi Kirsty, Very useful insight into how to recover from loss in SERPS. Goodluck and hope you manage to get it all back. Question how would you propose a Voucher site increase time on site, when all they are looking for is a code, once they have the code they are off to another site.

    I would also be interested to know what that plugin is you are referring to that grabs all the merchant images and resizes them. Care to share or e-mail me if you wish to.

    Thanks again for another great post

  6. Kirsty Says:

    Thanks Gary,

    With a voucher site that very question has been running through my mind! It’s a tough one. Perhaps other related resources about sales, merchant profiles etc. Bit lame is that I must confess – I think that’s a big challenge for voucher sites. Maybe some kind of tool to have people returning like a special offers saver where users can create accounts to save favourite deals or merchants??

    The plugin is one I’ve had created and will therefore not be available to the general public 😉


  7. Gary Says:

    Hi Kirsty,

    No problem worth asking hey and thanks for the tips on the voucher site. A big challenge you are right. My voucher site is like a dangling carrot. It keeps popping up in Google and then bouncing back down. When its on the 1st page its doing really well.

    Think I will write my own plugin to achieve what I need. Mainly grabbing the images as I’ve already sorted out the link masking, sizing of the images etc.

    Look forward to your next Panda Recovery installment


  8. Chloe Says:

    Hi Kirsty

    Thanks for taking the time to go into this in so much detail.

    A question though – please could you elaborate on what exactly you did to soften your SEO and what you are left with in terms of on page SEO?


  9. Simon Says:

    Kirsty – your tenacity in the face of another almighty google slapdown is really inspiring and your step by step recovery plan is like nothing else on the web. Thank you, it’s great to see you back blogging again. My lone affiliate site just got wiped out after it was just starting to make some decent money. It’s time I dusted myself off and got down to work…at least I know now which direction to take back up the mountain :–)

  10. Kirsty Says:

    Chloe – I’ve had it on good authority that Google is coming down on sites that obviously target a term aggressively. Signs of this include repetition of the same phrase in title, keywords, description, headers, and throughout text on page. If you take a more “natural” approach you wouldn’t necessarily see that in on page SEO. In short I think it comes back to creating pages for users, not search engines.

    Simon – thanks I’m glad you found it useful. Good luck in getting your site recovered. It is possible, just not the work of five minutes!

  11. Rob Barham Says:

    Hi Kirsty,
    Wow, amazing work. I have checked out your site and it works like a charm.
    Maybe you should sell a theme/plugin package for other affiliates 😉

  12. Kirsty Says:

    Thanks Rob, we’ve done a lot of work optimising page load times which seems to be helping us out a lot. I’m pleased with how it’s turned out!

  13. Mark Says:

    Thanks for your generosity in sharing this Kirsty.

    It takes a lot of guts to remove so much content from your site. I hope your efforts are rewarded.

    Did you delete your unwanted content or robots noindex it?

  14. Kirsty Says:

    Hi Mark, I deleted it as most of it really wouldn’t have been found / used by site visitors through natural use of the navigation. Some of it I may gradually re-work a bit and put back up – not all of it is totally dreadful.

    There was something very refreshing about starting with a clean sheet but it’ll likely take me months to build traffic back up again. Oh well… onwards and upwards eh?!

  15. Mark Says:

    Thanks for the answer above Kirsty…..

  16. Liam Delahunty Says:

    Looking at the long term stats image you have above it looks like your traffic dropped around January 14th. This isn’t therefore Panda as that rolled out Feb 24th and then again April 11th.

    Instead I think you were filtered/penalised, and the actions you’re taking are helping to alleviate that issue.

    Everything you’re doing is correct as regards the long term future of a quality site. I just wanted to clear up for any readers that this isn’t really mitigating the effects of Panda as, IMO, your site hasn’t been hit by it.

  17. Mark Says:

    Any chance of an update, please?

  18. Kirsty Says:

    Hi Liam, I’ll have to politely disagree with you there!! Although the issue did start in January, the site was also further hit on all the Panda dates. Hard to see on the graph there but definately a strong effect on remaining rankings.

    So Yep, the initial hit wasn’t Panda I must agree. but from my discussions with other site owners and also SEO’s there were some important pre-panda algo adjustments in the run up to the roll out. One in December and another in January. I think of it as leaving early to beat the crowd 😉

    Absolutely everything I am doing is tailor made to cope with the Panda algo (well I think so anyway!). My issues in January were, I think, due to adjustments in the over optimisation filter.

    If you have any further information on what actually WILL help people cope with the Panda update, please don’t hesitate to share it, myself and my readers would love to benefit from it.

    edit: For readers info. I do have a lot of info about boilerplating issues from voucher code type sites but didn’t share it here as I wanted to keep it simple. If there is such a thing when it comes to dealing with Google!

  19. Kirsty Says:

    Mark – not huge progress. I have gotten back a very large proportion of the key pages left on the site so they are at least now in the index.

    My strategy of binning such a huge amount of content never to return will probably mean it takes me months to rebuild the site to anything approaching it’s former traffic. However I believe it will be well worth it in the longer term.

    The traffic improvement I showed in the graph on this page has actually gone down a bit, but it’s now heading in the right direction again.

    I’ll do a comprehensive update once I have more information. I wanted to get to the next running of the Panda algo (which happened yesterday) to see how things were going.

    A couple of my other sites saw downturns yesterday but the core rankings (i.e the stuff that brings in 80% of the money) seem to be a-ok so far. I’ve not had a chance to fully investigate but I think what’s been hit are articles similar to the “thin” ones I deleted from the site above.

    Have you been affected Mark?

  20. Liam Delahunty Says:

    Hi Kirsty,

    I didn’t mean to sound negative about your post and please accept my apologies if it came off that way. I think you’re doing a lot to create a valuable resource and many of those actions are also appropriate for a site that’s suffered in the Panda update.

    Google’s Amit Singhal published a long list of questions to ask oneself about your sites here

    Regardless of Panda, that list is appropriate for building great websites. Taking steps against Panda is the same as building a better resource and both will result in better serps.

    One thing that I think is important is that you mentioned de-optimising pages accidentally due to a SEO plugin issue. De-optimisation of titles and other on-page elements is normally a quick fix against filters. You can then build a few more “branded” links and re-optimise the page(s) later.

    This is really semantics here, and regardless of if you got hit by Panda or if it was a filter, the changes being made are valuable and useful to both your sites and indeed to any webmaster.

    Against Panda I would tell a client, trim the excess pages, develop deeper content, build branded links to the home page and anchor links to deep pages; but frankly I’d pretty much give that advice to any site.

  21. Mark Says:

    Thankfully I’ve not been affected. After spending several years doing the wrong things, ie. the newbie-quick-win-approach and getting nowhere, I started doing the right things. I also focus on just one site rather than the 50+ that I was doing before.

  22. Kirsty Says:

    Hi Liam

    I didn’t take your comments negatively no!! I’m always pleased when anyone contributes to the blog – I was just wondering if you were saying that a different set of actions was more Panda appropriate or if there was a fundamental flaw / misconception in my strategy I couldn’t see. I long ago learned that thinking your way is 100% right is bloody stupid 😉

    Interestingly, I had two more sites affected in the recent Panda update. I lost about 25% of traffic, but oddly enough not in a negative way IYNWIM? I spent hours analysing what had gone and about 90% of my core pages and “money” terms were still ranking fine. What seems to have been stripped away are a majority of “accidental” rankings, where you see a search landing on a page and think “How the hell can that be relevant to XYZ??”. Fingers crossed it stays that way – I’m working hard on both those sites to do the same as the site above but without detonating absolutely all content outwith the site core.

    Thanks for the extra information too – I do need to build links to the core pages. Interestingly, the pages that have recovered most strongly are those where a small amount of work of this nature has already been carried out.

    Any thoughts on what the recent re-run was designed to do? Disappointingly, there’s still loads of scrapers ranking. Most frustratingly, some with the old content that Google refused to rank from my site. Gahhh.

    Thanks again for the contribution – I hope you’ll pop back again for more Panda and other SEO type fun, hahaha!

  23. Kirsty Says:

    Mark – really glad to hear you are unaffected! I think a lot of “lazy” affiliates will struggle to recover their slapped sites.

    I think having large sites with a lot of “old” content probably doesn’t help me. I’ve been going through and deleting off anything with shocking bounce rates and more than 18 months to 2 years old. Annoyingly, I have a friend who has sites very similarly structured to mine who has been totally unaffected – the only difference I can see is that his link profile may be slightly better, and the content is a lot newer. Confusing!

  24. Liam Delahunty Says:

    “Any thoughts on what the recent re-run was designed to do? Disappointingly, there’s still loads of scrapers ranking. Most frustratingly, some with the old content that Google refused to rank from my site. Gahhh.”

    Google’s Moultano has some interesting points here:

    “I’ve never seen an actual instance of a site with original content being penalized because it is getting scraped (though it is theoretically possible.) Our systems for this are robust and quite conservative.”

    Though that appears contrary to what I’ve experienced. He continues:
    “When a scraper outranks the original site it’s because we weren’t aggressive enough in demoting the scraper, or don’t have enough data about it, not because the original was penalized.”

    So, more reports to the web spam team are probably in order.

    As regards the latest Panda which some are calling 2.2, I haven’t seen the effects on my sites. Someone at Google (Matt Cutts?) did say that Panda wasn’t real time and would they have to re generate the data from time to time.

  25. Kirsty Says:

    Thanks Liam, much appreciated. Interesting stuff there! And yep, it was Matt Cutts who said Panda would be re-run manually.

    I’m still struggling to understand the latest update and why it’s affected certain parts of one of my sites. The really odd thing is that whilst referrals have dropped off about 25 – 30% the actual page views are the same, and bounce rate has dropped dramatically. In terms of earnings there’s no real drop.

    I think I’m going to go away now and have a quiet migraine for an hour or two. I guess if you do lose a slice of traffic this is the stuff you’d want to get rid of. Still a worry when you don’t understand why 😉

  26. Claudiu Says:

    Hi Kirsty, did you see anything change after July 23rd (latest panda update)? We had been hit hard by panda too, but made a recovery in July, wrote about it here – sort of an incremental change, from 70% traffic lost to only 20% lost, but still something positive.

  27. Mark Says:

    Just thought I’d add a few things I’m doing to combat a recent drop in traffic (-90% .. ouch!). I’ve added the excellent W3 Total Cache plugin to help speed up my site. Average load times have gone from 13-15s to less than 5s. This, in part, has also been due to the addition of a CDN.

    I’m also in the process of removing every affiliate link from the page. Rather than using a cloaking plugin, I’m redirecting them via a PHP file.

    Finally, like Kirsty, I’ve deleted all poor quality content.

Leave a Reply

Subscribe without commenting

 © Copyright 2010. All rights reserved