PANDA International - Data on Winners & Losers

[tedster]

This afternoon the SearchMetrics blog released their data on the international wave of the Panda Update - affecting English Language search. The article includes data for 100 losers and 20 winners, along with some dramatic graphs.

The data was collated between the 5th & 12th of April 2011 and we can definitely confirm that the update has hit the UK – in a big way.

Surprising is that ehow.co.uk and ehow.com has lost more than 50% in visibility and the alarm bells are probably going off at Qype.co.uk who’ve lost a whopping 96%! A lot of price comparison sites like ciao.co.uk (in a lawsuite with Google) and dooyoo.co.uk also lost nearly 90% of visibility. [blog.searchmetrics.com...]

This thread is for discussing the data - if you wish to share editorial opinion, positive or critical, please post in our other thread: Google PANDA rolls out WorldWide [webmasterworld.com]

[johnhh]

A few of Doug's competitors are saying the same over at a well known UK affiliate forum ( where Doug also posts )

Mind you, he has a massive mail list and probably will never be affected that much by Google rankings.

<edit>for grammar</edit>

[chrisfurther]

Thanks tedster. We have some people commenting and retweeting saying they know other sites on the list that have not seen drops. I'm starting to question the figures, although I wonder why more companies haven't come forward publically?

[binomsc]

Just one small piece of information which may help to better understand the figures. Both, Sistrix's as well as Searchmetrics'numbers, especially their KPI's (key performance indicators) which are named VisibilityIndex (Sistrix) or OPI - Organic Performance Index, are not directly related to traffic.

Both KPI's try to give an idea about a website's presences ("visibility") in the SERPs. Basically they use something like "number of keywords the website is listed" weighted by the "average keyword position".

Sistrix explains it shortly at the blog post [sistrix.com] "The SISTRIX VisibilityIndex is an index value calculated from traffic on keywords, ranking and click-through rate on specific positions."

Simply put: a website which is listed for many keyword (combinations) on a high position on Google will be seen (and thus clicked!) more often than a site who is only on page 5 of the search engine results for a couple of search phrases only.

[binomsc]

@danimalSK:

These numbers are pretty far off. I know first hand that a couple of the sites they are reporting at 80 to 90% "visibility" loss are down about 30% in actual traffic.

Pls have a look at my last post to understand what the OPI and the VisibilityIndex try to measure. BTW both, Searchmetrics and Sistrix, have similiar/same webpages in their top looser list (of course not in exactly the same order as they use different data sets). I guess it is important to understand the metrics, what they express - and what not.

Additionally it is important to understand that both companies monitor a huge set of keywords (millions), however not every long tail keyword. Could it be that the sites you have in mind are quite active in the long long long tail? It might be that they still are in the listings for those 3+ word search phrases.

I try to make a small example: I assume that the "couple of websites you know first hand" have more traffic sources than just organic traffic from Google.co.uk? Let's assume as a simple example they have 100 visits/day, 50% from Search Engines Organic. To keep it really simple: let's assume it's only Google.co.uk - that makes 50 visits. The others from direct access, or image search, BING, or affiliate, backlinks/referrals, you name it.

Now we do some math: assuming the site loses 90% of visibility in the SERPs, it loses approx. 90% of it's organic Google.co.uk traffic, which are 45 visits. So they still have 55 visits in total, a loss of "only" 45% in traffic.

Now you can play with the numbers (percentage of organic SE traffic of "your" sites, part of "long long longtail" keywords which are not considered in boths KPI's, click-through rate of the remaining keyword listings etc.). I wouldn't be surprised if this explains why it is "only" 30% traffic loss for the sites you mentioned.

[danimalSK]

Erm, dude, no offence but I have a degree in engineering, an MBA, and used to work on stats models for one of the biggest investment banks in the world. I don't need a lesson in basic percentages.

Sistrix and Searchmetrics aren't anywhere close to measuring enough keywords to meaningfully cover all these sites. I know because I've worked on some of them. A site like Qype gets around 500k unique keyphrases hit it a month, and thats just ONE site on the list. They are very very long tail.

[binomsc]

Okay, great! Maybe I did just misunderstand your posting, I apologize. I am not a native speaker.

For me it sounded a bit like you were expecting a 90% loss in traffic. At least that was my interpretation of your short sentence I quoted. And just in case some other had the same interpretation, the little example might help them ;)

Regarding the number of keywords they used for their research:

"Like last time, the following table is based on a dataset of one million keywords, which were checked before the update and yesterday/today after Panda went live." (from Sistrix article) [sistrix.com]

We have historical data for around 8 million UK keywords for the last 10 months. (Marcus from Searchmetrics, commenting on a blog post) [seroundtable.com]

Not sure if that is enough or not, for me it sounds like a pretty huge sample. At least a start for analysis, and as far as I remember my professor for statistics: One has always be aware of the limitations of ANY statistic. Again, just my two cents.

[chrisfurther]

UPDATE: For anyone interested The owner of discountvouchers.co.uk has actually sent us a "screenshot" of his Google GA organic traffic. Small dip but certainly nowhere near 60% what was quoted by search metrics. The image is on our post here: [further.co.uk...]

[dougs]

If anyone wished for me to send them an image of our analytics then please contact me. Our traffic levels have not changed.

I am pleased to say that our staff and customers are no longer worried. After Search Metrics published some false data I have had to calm fears. Maybe SearchMetrics data is not what they are stating.... Check your facts guys.

Regards

Doug Scott
MD
www.discountvouchers.co.uk

[Leosghost]

Some of us did express reservations regarding that "data" from Search Metrics ..interesting to hear from the "horses mouth" ..if you'll pardon the expression Doug .;-)

[browsee]

Just checking hubpages in quantcast(directly measured), they are way down. SearchMetrics is not wrong with hubpages.

[walkman]

I am pleased to say that our staff and customers are no longer worried. After Search Metrics published some false data I have had to calm fears. Maybe SearchMetrics data is not what they are stating.... Check your facts guys.

Hehehehe, dude you are pushing your luck by posting all over about this. Google engineers read things and there is no double jeopardy with Panda.

[potentialgeek]

I have about nine sites which were hit by Panda 2.0 that weren't hit by Panda 1.0. I've been looking over the ones which were hit and others which weren't to try and figure out what the deal is.

The thing I find particularly annoying about this algo update is it completely ignores site age. The demoted sites were built 5-10 years ago and didn't have any ads for years after they were built unlike classic content farms.

It also ignores ad network reviews by the ad network. I get that Google AdSense only requires one site to pass a test when the publisher first joins, and then it stops reviewing, but that's not how it works with reputable ad networks. They scrutinize each site you want to add. They won't allow junk, because the advertisers will complain.

It looks as if Google is becoming Yahoo (1995). A directory of brands. How cutting edge! This is really lame and retro. It basically is the "easy way out" and requires no intelligent engineering. It's an admission of failure. "We can't figure out how to tell good sites from bad so let's just go with brands."

Hitting an entire site because of one bad page is extremely heavy-handed and thuggish. It's like Gaddafi murdering an entire family because one family member was a rebel. Once again, it's a very weak attempt at engineering. A site can have 100 great pages but one bad one -- which nobody even visits -- but Google declares it a bad site.

Whatever happened to "Don't be evil"? It's the heavy-handed actions which alienate people from companies, leaders, etc. Panda is a huge brand mistake for Google.

[dougs]

My reason for posting was simply to make sure our suppliers and staff were aware of what was really happening.

Personally I love what Google are aiming to do, as it will reward sites that try to offer unique content. We know the issue in the discount voucher business in that many people get the same vouchers from feeds, the same as we do, but also we have a full team of people doing deals and hand finding deals and vouchers that no one else has.

On any one day we now have upto 3500 deals, and we are just about to completely restructure out site to give a better experience. The speed at which we have grown has come as a great surprise and sadly the site has just grown into an unwieldy mess:) Over the next few weeks our new design should start to rollout and hopefully the new Panda world will love us more.

Doug

[rusj407]

Still trying to figure out Panda 2 as I have found that in the top five results for various keywords there is little consistency as what seems to be working.

While I do see less Adsense bloated sites, the number 1 site for two top keywords where we were in the Top 5 now has pink on pink hidden text (seriously that is so HotBot, ughhh).

#3 Site had over 4 Adsense and less than hundred words of text on the page.

The part that does piss me off is that I am seeing the several sites in the top 30 that are different urls with the same graphics, same link blocks but different content.

So I am a bit frustrated that my site that I typed word for word and added Adsense to because the Adsense Reps called me to do it, may have played a downfall in my rankings.

One thing I did glean from reviewing top 30 for several groups of keywords was that Google is looking for natural sentences on the topic of the page that is unique.

So I am going to remove some Adsense units, Add an Anchoring Sentence or two and gauge the changes over the next week.

Russ

[Whitey]

Has anyone who tanked on Panda in the US seen a traffic rebound since ? I'm hearing unconfirmed reports that this has occurred on some sites.

[Shatner]

@Whitey What sites have you heard rumored?

The only recoveries I've heard of have come as a result of Google basically whitelisting people (whether they want to admit it or not).

[walkman]

Apparently sistrix and others are useless, even eHow is disputing their findings.

Certain third parties that have published reports attempting to estimate the effect of recent search engine algorithm changes made by Google on traffic to the Company’s owned and operated websites have significantly overstated the negative impact of those changes on traffic to eHow.com, as compared to the Company’s directly measured internal data.

Recent search engine algorithm changes have negatively impacted search driven traffic to some of our websites, including eHow.com, resulting in moderately lower year-to-date page view growth for the Company’s owned and operated Content & Media properties compared to page view growth rates before the algorithm changes.

Nevertheless, the Company currently expects that its year-over-year page view growth across its owned and operated Content & Media properties in the second quarter of 2011 will be comparable to, or greater than, the year-over-year page view growth achieved in the second quarter of 2010.

As I said in my prior post, we generally do not comment or speculate on changes by major search engines, as these changes can happen nearly daily. However, recent third-party reports attempting to estimate the impact to our search driven traffic, including one projecting a 2/3rds decline in eHow.com traffic, are so significantly overstated that we decided to comment.

As discussed in our press release issued today, we currently expect that in Q2 2011 our owned and operated Content & Media properties will generate year-over-year page view growth comparable to or greater than the year-over-year page view growth reported for Q2 2010. We have also reaffirmed our calendar year 2011 financial guidance in this press release.

[kara.allthingsd.com...]

[tedster]

I trust the data in one way - as a metric for gain or loss in total keywords that are ranking. Hoping to extrapolate from there to rankings or even earnings is a dicey game. Especially for publicly held companies, this may increasingly become a PR battle to control the message.

But for working toward the factors that might be included in Panda's assessment, both company's data does seem potentially useful to me. However, as we've noticed, clear answers are not falling out very easily.

[edited by: tedster at 2:51 pm (utc) on Apr 18, 2011]

[brotherhood of LAN]

Yeah I think their data(set) is

- useful for measuring a general trend, providing the keyword coverage gets the major search terms of the niche and its not a long long tail area of search

- changes within a search terms rankings

- good coverage of niches where a handful of terms cover the majority of searches

A 'global' report like the lists produced have to rely on having comprehensive data for pretty much all the sites that make the top lists, a tall order. So, at least (globally) they are indicative of movement but less reliable for traffic measurement.

* 20% of Google queries not seen before in past 3 months
* 64% of Google searches have pages without exact matches to all query terms
* 44% searches on google have more than 3 words

This was mentioned on Danny Sullivan's twitter via Matt Cutts.

[enigma1]

Here are some of the more well-known sites besides those already mentioned that the metrics firm says got demoted:

about.com -28.58%
.......

This is from a post on the first page of the thread. To tell you the truth I am not surprised about these numbers, as some sites listed there - especially the first one, says one thing and does another.

I believe the general populous will gain more than lose anything from the particular demotions by google.

[tedster]

Thanks to browsee for an alert on new data from Experian Hitwise. As reported by Forbes blogger Jeff Bercovici [blogs.forbes.com], the focus of this study is specifically on Demand Media's web properties. It appears that they have lost a lot of Google traffic with this new piece of the Panda Update.

Hitwise looked at downstream traffic from Google — ie. what sites do users surf to next after visiting Google.com. In the first two weeks of January, 0.57 percent of those who departed Google next visited a site operated by Demand Media, the best known of the content farms. That proportion hovered at 0.55 percent through the last week of February, when the initial Panda changes took effect.

Those didn't have much impact on Demand's Google referrals, which subsided only slightly. But by mid-April, with the full suite of Panda updates in place, Demand was feeling the pain. As of April 16, it accounted for only 0.34 percent of Google's downstream, a 40 percent decline from the start of 2011.

Article [blogs.forbes.com]

Here's the URL for the related Googledocs spreadsheet:
[spreadsheets.google.com...]