What Happend To My Ranks? Began Dec 22 - detailed Info Inside

[mattzer]

I have a .co.uk domain that was setup in November 2009 and it has always done pretty will in the SERPS. In Oct-Dec last year I could literally publish a post and be on the first page for a keyword with no backlinks at all.

All of my posts/pages feature-

100% unique content (min 500 words)
Imagery
Embeded video which is part of a video sitemap
External links
keyword density between 2 and 4
keyword in first paragraph bolded
Title tags are written like KEYWORD - Title, eg BLUE WIDGETS - What You Need To Know

Then this happened-

[i53.tinypic.com...]


On december 22nd - december 24th i see a sharp decline in traffic, and then a further decline december 31st - Jan 1st?

On that graph I have marked out changes I have made to the site. The bottom bar that was deployed on Nov 30th was simply a floating price comparison bar stuck to the bottom of pages.

Other than that change, nothing else had changed on the site during that period.

Here's what I have done so far to try and fix things-

Optimized load time (now 2.5 seconds)
Deployed MaxCDN content delivery network
Removed some pages which were similar, but 100% unique (unlock pages on graph)
Change H2 tags to H1 (previously there were no H1 tags)
Removed the bottom bar on nov 30th
reduced price compare results to 3 from 10
Ran a broken link check
Noindex/Follow 2nd pages of tag/categories pages
Improved blog posting rates to at least twice a day

I have 100% indexation rate across the site, an average load time of 2.5 seconds. I've tried everything I can think of, and can't seem to pin point the issue and after 5 months of trying to get it fixed I think an outside view is needed!

So what happened at Google mid-end december?

Any advice/thoughts/feedback is greatly appreciated, thanks for your time.

[tedster]

This is a situation where our monthly SERP Changes thread might come in useful. Here's the thread that applies: Google Updates and SERP Changes - December 2010 [webmasterworld.com]. And here's the historical index of SERP change discussions: [webmasterworld.com...]

Scanning back toward the end of that month, I see a discussion where several people noticed more results from the same domain showing up in the SERP all of a sudden. There also seemed to be a leap in personalization being used for more query terms, and a shift for some sites to ranking an internal page where they used to rank the home page for the same term.

So on Dec 22, 2010 there was an algorithm tweak of some kind. Your site might have taken the brunt of that, getting pushed down the page or even to page 2 for certain terms. If that is what happened, then another site or group of sites were seen as the best result for the intention behind the query - and they got multiple rankings.

The second SERPs tweak you saw, right over New Year's Day, slipped through our radar here - no one posted about it - one report of a traffic spike from one member, but no analysis of what changed.

[edited by: tedster at 5:54 pm (utc) on May 14, 2011]

[anallawalla]

In Australia, we see drops in traffic on Xmas Eve, Xmas/Boxing Day, and the New Year's Day period. Does your site not have variations on public holidays?

As for being on page 1 for any post of yours. Was that the case if you checked that post a month later?

[Planet13]

In Australia, we see drops in traffic on Xmas Eve, Xmas/Boxing Day, and the New Year's Day period. Does your site not have variations on public holidays?

Looking at the traffic graphic he posted, the traffic never recovered after the drops around Christmas time.

[Robert Charlton]

There are several gaps in the information about the site and the graph that need filling in to help with interpretation....

Is the traffic UK and US, or UK only? I ask because if US and UK combined, then the drop very soon after the USA Panda might be attributable to US Panda. If that's UK only traffic, then it's not likely to have been the US Panda. Traffic actually seems to increase for a short while with UK Panda.

It's also not clear whether your traffic is seasonal, or what might be expected on holidays... whether searchers are more likely to search from home or at work, etc. I'm tending to agree with anallawalla's thought that the big holiday dips in traffic were most likely due to the holidays.

If you smooth out your curve to adjust for the holiday weekends, though, then you've got a noticeable overall downtrend starting around Dec 6 or 7th, with some additional steep drops for various site changes and for US Panda.

I get the feeling that there's a lot that's technically messed up on the site. You mention a 2.5 second loading time in this thread, but on a late January thread [webmasterworld.com...] (which I assume is about the same site), one of your graphs [i55.tinypic.com...] shows an average loading time of 9.8 seconds, "slower than 94% of sites". The big mid-January rise in load times in that graph more or less corresponds to your big mid-January drop in visitors on your current traffic graph.

I think you need to give us (and yourself) a very clear time line. It's difficult to jump around and to read different interpretations of data, especially when some data is in graphics and some posted in the threads in text, and not chronologically, and over two threads.

You might also want to recheck some of your information, as some in this thread, eg, appears to be inconsistent or confusing, as in....

The bottom bar that was deployed on Nov 30th was simply a floating price comparison bar stuck to the bottom of pages.

Removed the bottom bar on nov 30th
reduced price compare results to 3 from 10

Hard to tell what's going on from that... whether whatever the changes were could have precipitated the early December drop.

No doubt all this plays into Panda too.

[mattzer]

Hi,

Thanks for the feedback guys, I can't work out how to quote posts so I will reply individually below-

Tedster, thank you for the link to that thread, I had read through it previously and there was nothing that really struck me but is certainly worth reading over again. Is there a tool I could use that would be able to show me the top 10 ranks across those days in December? I don't record that kind of data but im sure there is some kind of tool/service that does.

Anallawalla & Planet 13, The site isn't based on seasonal traffic at all and targets product specific terms.

Robert Charlton, The site is made up of around 90% UK traffic, but it is strange that the drop coincides with usa panda.

The site isn't seasonal but based on electornic products targeting by product name.

The post you reference is about the same site, however since then I have made the list of changes in the first post and spent significant time optimizing my server, wordpress setup, image compression and the use of a CDN.

Sorry about the confusion of the bottom bar, that was ment to say I removed the bottom bar that was placed on Nov 30th.

What I have done since writing this post-

Move the header code into footer so the content is the first thing in source
remove wordpress comments for post feed from header
moved twitter share and facebook like buttons to end of post

What I also plan to do-

Currently my titles are in the following format -

#KEYWORD# >> What You Need To Know

And my H1 tag is

#KEYWORD#

I believe the title and h1 tags maybe over optimized, I would be interested to hear others peoples opinions on this.

What I plan to do is rewrite all of the titles so they appear like a proper sentence and then also make the title the H1 tag, eg-

What You Need To Know About #Keyword#

I should also note that usually after the first paragraph of content I display a price comparison table with no more than 3 links displayed. The affiliate url's are masked with a 301 of the same domain.

Although it doesn't appear that UK panda had an impact on the site, it did include the placement of ads in the algo. I'm not sure if Google would identify my price comparison table as an Ad or not. What are your thoughts around this?

Thanks for all of the input and taking the time to understand my problems.

[tedster]

Is there a tool I could use that would be able to show me the top 10 ranks across those days in December? I don't record that kind of data but im sure there is some kind of tool/service that does.

Not online and for free, that's for sure. The scale of data archiving that would be needed is immense. Rhe companies who provided early Panda analysis like Systrix and SearchMetrics do some of that for a limited keyword set, but it takes money to get the data.

I think we usually do best to operate in present time anyway - who is now on the top SERPs and what do we need to do to match that performance, all around.

[Planet13]

I have some further questions, if you don't mind, as I think it could help us get a better idea of your site:

Firstly, is it safe to assume that your site is a review site with reviews written by yourself?

How important is user-generated content to your site?

Are the images you use distributed by the manufacturer or some other source and are they the same images seen on many other sites?

Are the embedded videos unique to your site? Or are there several other sites using the same videos?

What keywords were you ranking well for previously? Were they following the format of your title (i.e., Keyqord what yoiu need to know)?

Or were you ranking well for just the single "keyword"

What sites are now ranking above you in the serps? What do their titles read like?

[tedster]

What keywords were you ranking well for previously...
What sites are now ranking above you...

Because mattzer is new here, I want to clarify that those questions cannot be answered literally in our public forums - no specific keywords or websites. But if he can paint a kind of picture for us, it may hep a lot. (E.g. "I was ranking for product brand names, or model numbers, or words that are generic product description." "I am now outranked by scraped versions of my own content, or major corporate websites, or other sites that do the same thing I do.")

[Planet13]

Thanks for pointing that out, tedster. That was what I meant to ask, and having been on these forums too long, I assumed he / she would know what I meant.

[mattzer]

Hi guys,

Thanks for your continued efforts :)

Planet13, the site reviews products in the consumer electronic market which is supported by price comparisons via a datafeed backend. It's all done in wordpress.

User generated content comes in the form of comments on the posts but there are no user generated reviews/ratings etc

The images are from the manufactuer but these are always resized and watermarked. I usually put them in a folder structure like this www.domain.com/wp-content/gallery/KEYWORD/KEYWORD.jpg

The embedded videos are completly unique to my site

I'm a bit confuised with what you mean here 'What keywords were you ranking well for previously? Were they following the format of your title (i.e., Keyqord what yoiu need to know)? '

I wouldn't mind pm'ing my site to someone if that would help. I understand its difficult to build a picture from forum posts!

Thank you all again for your help with this!

[anand84]

I am sure you have considered this. But for sake of clarity, is the drop in traffic distributed across several pages or was it the case that 5-6 high traffic pages dropped out on the SERPS? Also, where are you ranking now?

I am wondering if it is possible that your recent website updates have simply pushed your high traffic pages deeper that have could have affected its ranking.

[TheMadScientist]

I just skimmed through this and haven't seen it mentioned, so if it has been my apologies, but isn't the timing of the drop about the time Google announced they changed the way they scored backlinks?

I think there were a couple of announcements late last year and early this year about backlink scoring changes and the date of the drop seems to be right around those time periods, if I remember correctly.

[Rlilly]

It could very well be seasonal. For the lasst 8 years, every Dec 19 to 22nd we have a huge dip in traffic. It slowely comes back around March.

[mattzer]

Quick update, all my rankings dropped out the top 500 Monday morning - Tuesday evening when they returned.

Anand84, I suppose the majority of traffic (70%) does come from 5-6 pages, however these were part of the sidewide rankings drop.

TheMadScientist, I would like to hear more about this as it would explain the sudden drop in December. After much reading I had come to the conclusion that there was no significant change at Google during that time period so thats music to my ears :)

Rlilly, I dont believe it to be seasonal. Although I do expect drops around xmas and new year.


I have also done some further analysis,

Firstly I analysed the traffic categorys for the past 8 months based on this post [seomoz.org...]

Which produced the following graphs-

All traffic on site - [i53.tinypic.com...]
UK organic search traffic only - [i53.tinypic.com...]

I then took 4 keywords for my niche, and analysed onsite and offsite factors for the top 3-

[spreadsheets.google.com...]

[learnseo81]

@mattzer

You are saying that all your rankings dropped on Monday Morning and on Tuesday Evening they all returned?

You've got the same rankings or better? Just curious to know.

[mattzer]

Hi LearnSEO81,

Yes exactly that, they returned to what they were before the drop on Monday.

[TheMadScientist]

Here's a couple threads for you:

Reason for Dec.28th Ranking Drop: [webmasterworld.com...]

Profile Links: [webmasterworld.com...]

I can't find the specific post(s) talking about backlink scoring changes the end of last year / early this year, I think it was in one of tedster's posts where I remember seeing a mention of a couple of announcements, but I'm not 100% sure ... It was one of those things I 'filed away' as 'good to know', but I don't remember exactly where it came from! lol

Sorry I can't be more specific, but I'm about certain there was a 'Pre-Panda' message about a couple of backlink scoring changes between late last year and late Jan. this year.

If you want to have some fun and try to find a reference, click the search link at the top of the page and try the one that says it's offline ;)