I promised I would keep a bit of a log on this blog about the process I’m going through to try and sort out some Google filter issues my lingerie website has been suffering from. I’ve now spent almost a week working my way though the site and have altered about two thirds of the content. As I’ve gone along I’ve made loads of test searches which helped focus on the issues that were present.
Some interesting developments have happened in the last few days. I’ve got screenshots of this, so I thought it was time for another post!
My (Slightly Revised) Theories on The Causes
As I worked my way through the site I discovered a distinct pattern to those pages that were still ranking – they had 100% unique content on them. These would have been the pages that immediately bounced back into the index after my initial problems. I found that loads of the pages filtered out had inadvertently used the same or very similar content via a combination of the blurb I place on each page and similar post excerpts that were also on many other pages.
However, I am now sure there was a second filter being applied… perhaps one for keyword stuffing / over optimisation. If it had been entirely dupe content, I think I’d have seen a more even spread of Google “picking” inappropriate pages to rank. However, a number of pages were also behaving in a different way and were either not in sight in the index or were ranking Waaaaaaaaayyyyyyy back.
There’s also the small matter not mentioned previously – I have a second site affected which has no duplicate content. It’s just been shoved right back in the results as per some of the lingerie pages, and shares the same site structure. This is what convinced me two things were happening on my lingerie site.
So Here’s What I Did
I am visiting each and every page on my site and: -
- Removing the similar output from the top of every page and replacing it with a paragraph of unique text which then links to the brand description further down the page via a named anchor (not an optimised one I might add!).
- Making sure that each and every brand description is completely unique.
- Removing any pages that don’t need to be there. i.e. if one page can cover the swimwear and lingerie lines for a specific brand, I’m consolidating my content.
- Manually altering each and every similar post output so that each post excerpt is only ever featured on one brand page.
- Removing any unnatural looking repetitions of my target keywords in the body text.
- Reducing the amount of header and anchor text that mentions target keywords.
Now For The Interesting Bit. What Effect Has This Had?
Well, it’s still early days as Google’s cache for the pages I’ve altered has still not updated. And yet I’ve started to see something very odd over the last couple of days. Something that MIGHT back up my keyword stuffing / over optimisation theory OR could just be a Google funny that has coincidentally happened this week. I can see other sites showing similar symptoms today including ASOS… so I’m not totally sure this is related to my filter issues.
Some Pages Have Started To Rank Again – But Not All The Time… and Still Not “Properly”
This is really fascinating actually. I first noticed this yesterday and have been grabbing a few screenshots. I know it was not happening prior to my changes as I have analysed each and every Google referral I’ve had for the last two weeks!
OK. Here’s my site ranking for “Manstore Mens Underwear“. This page was not ranking at all a couple of days ago, regardless of which search string used.
As you can see though, Google is not showing the title in bold which I’ve never seen or heard of before (which makes me think it may not be entirely penalty related – but it sure is interesting!) The same page also ranks for “Manstore robot strip string” and “Manstore Underwear”
*update- I am now seeing this “title not in bold” issue manifesting itself all over the place – so its clearly Google fiddling around generally that is causing this rather than any kind of penalty.
For the search term “Manstore Boxers” which you could logically assume I should be ranking for, this is the page Google has chosen to place at result 50: -
*update – this has changed over the weekend. I’m now ranking number one for the term “Manstore Boxers”
I’m seeing this time and time again over many pages. Some suffer from the “Non Bold” issue I’ve described above, others are fine. Here’s another interesting example that made me think a keyword specific penalty has been in place on the following page alongside a dupe content penalty (which again had previously been knocked out of the rankings altogether).
Here’s the page being ranked for “DKNY Nightwear” at result 10 (hurahh!)
But here is the search result for “DKNY Sleepwear” waaayyy back at result 82
I’m seeing the same on a Calvin Klein page I altered a few days ago. I’m ranking 15 for “Calvin Klein Nightwear” and again a “wrong page” is popping up for “CK Sleepwear”
*edit - I’m now seeing the right page ranking for CK Sleepwear, around result 35. Progress!
I’m seeing similar on other pages, but at the moment it is in a constant state of flux. Some that I checked right before writing this article had changed by the time I got to the end here, with the “right page” popping back up. So that makes me think: -
- Some of my pages are being slowly forgiven.
- Some of the pages are exhibiting a penalty for particular search phrases i.e. those containing “Sleepwear” because the page content was too tightly aligned with them.
- I’m going totally and utterly mad and its time to move into the “Sunnyvale Home for Knackered Old Affiliates”.
Finally….
Well, that wraps up my thoughts for now. I hope it is helpful to anyone who is interested in this kind of stuff. This on its own is great fun actually. However the other site thats just been bumped to the end of the results worries me a fair bit, there’s much less to go on with it. All I’ve learned here will be a good place to start though.
Wish me luck 
August 8th, 2008 at 5:47 am
Interesting results. I had never seen that kind of duplicate content penalty for blurbs before. Certainly plenty of news sites do pretty much the same thing but you never know. Perhaps affiliate sites are ranked by a different part of the algorithm. I had never really heard of “over-optimization” either except in the case of blatant keyword stuffing but perhaps you used the keywords a wee bit too much. I suppose it’s a fine line to tread. We’ve certainly been having good results with some of our client sites in certain markets that have terrible on page content but good page titles. Yet, others in more competitive markets have been taking some hits but we think it’s got more to do with hidden text in the CSS that, even though it isn’t hiding keywords, is still getting penalized with the old display: none CSS property. As a general policy, we’re trying to steer away from using it in our designs. It’s a shame too because it’s a great technique. Anyway, I’m enjoying your experiment.
August 8th, 2008 at 8:05 am
Another great post Kirsty. There are definately a few sets of very different data kicking around Google at the moment. From what I can see there has been some significant changes recently and that data is still filtering through.
August 8th, 2008 at 8:32 am
Don’t know what to tell you Kirsty as I’ve had this problem for quite a while with some of my pages – won’t go into specifics – BUT I can tell you I’ve agonized for hours and hours, tried everything and still can’t get the correct page to rank for the correct term – it’s ludicrous. I’ve spent far too much time (I also find it fascinating) to the detriment of other area’s.
BUT the one thing that has worked really well is, what I like to call ‘waffling’ when writing a description – picked up tons and tons of long tail phrases, which is reflected in my sales, which have done rather nicely, especially this week – if you want further details PM me.
August 9th, 2008 at 2:51 pm
Here’s an idea — sit back and do absolutely nothing!
Sounds pretty stupid, I know. But Google seems to be in a state of flux right now. Reacting too quickly to these things can just send you round in circles. If I notice some significant ranking changes, then I try (really hard!) not to touch anything for a couple of weeks to let it settle. Quite often things sort themselves out without me doing anything. Simply sitting back and observing will also give you an indication of how much Google is still shuffling things around. If you react too quickly and start making major changes, then it’s really difficult to make any sense of what’s going on. It’s very difficult to leave things alone when you’re losing sales though, I know! Also, I seem to remember Matt Cutts saying that some old filters were going to be lifted (can’t remember how long ago this was though). So it could have something to do with that.
August 10th, 2008 at 10:50 am
Hi Kirsty interesting stuff…I was just wondering how the duplicate content pages affected your index page?
Did you see an increased ranking due to your fixing of the problems on the subpages?
August 11th, 2008 at 4:09 am
Thanks for the comments and thoughts everyone!
@ Chromate – yeah I did try the old “do nothing” technique for a few weeks. However, now that the issue is around 3 months old I think I need to take action. The stuff that I’m doing is really “obvious” type things. Even if they aren’t causing me problems now they might later. A good reason to try and sort them out!
@ Mobilica. The index page has always ranked really rather well since the first set of issues was sorted out (duplicate URL’s).
August 12th, 2008 at 2:42 pm
I think Google is definately altering a lot at the minute, did you run your pages through to see if your keyword density is at a good level?