I remember, at the tender, impressionable age of 11, my mathematics teacher asking the class how many point you needed to draw a line. The hands shot up and the answer 2 was duly given. The teacher said that while that was correct, in her opinion you needed at least 3 points to truly know whether the data truly indicated a line.
I’m sure that most people reading this are aware of the Panda 2.5 update that ran last Wednesday. It first came to light when some sites noticed their traffic dropping dramatically, and was confirmed by both Google and Google’s Matt Cutts.
Different sites, this one included, jumped over to SearchMetrics and used their SEO visibility chart to identify the winners and losers of the latest iteration of Panda. The data seemed to show that Google was now favoring their own brands, because they were the ones that had seen the biggest jump in this “SEO visibility” metric that was being used to identify Panda impact.
What Exactly is SEO Visibility?
I looked on the SearchMetrics site and found this definition:
“SEO Visibility displays the visibility ot the selected domain in the organic search”
Not an exceptionally helpful definition, but OK, let’s just roll with it, and assume it’s a valid metric that holds up for all types of sites including those in the news industry with slightly more volatile rankings due to the news cycle (i.e., 2 weeks ago sites weren’t ranking for “Steve Jobs Dead” or “Amish on Amish Crime” because they weren’t news yet).
A Closer Look at the Panda Winners
FoxNews.com was highlighted by SearchMetrics as a winner on their Panda 2.5 blog post based on their increase in SEO visibility between the week before and the week after Panda 2.5 ran. Congratulations Mr. Murdoch. Oh but wait, what if we look at more than those two data points? What does that show?
Hmm, while this metric does indeed show an increase for FoxNews.com between those two weeks, it also shows that it had been on a downward trend before that, and even after the increase purportedly due to Panda, it was still down from where it had been three weeks earlier. Then a week later... back down to almost the low of the week before Panda. Not really what I’d call conclusive proof of FoxNews.com being a Panda winner.
Maybe that’s an isolated case?
How about a Google owned site? Because this update allegedly benefited Google’s products. Android.com was listed as a 10% gainer. Further proof of Google sticking it to sites they don’t own by helping their own? Let’s look at the chart to see if that’s actually the case.
Well, yes, that increase was there, but it was a continuation of a trend that began the week before, and in the last week... it’s dropped again. So looking at this data, you could say that it shows a 6.2 percent increase over the five weeks, or you could say that based on the visibility score now compared to the week before Panda it’s actually fallen by 0.8 percent, not much of an own brand favor.
How About the Panda Losers?
Well, the losers were identified according to two data points.
Consumer Affairs had a 78 percent drop... that is until the next week when it had a 437 percent increase.
Maybe they saw the drop, and fixed their issues in time for the next data cycle? Let’s give that one the benefit of the doubt and look at a few others. Hmm, the same “drop and improve” pattern is true of Businesswire.com, starpulse.com, savings.com, PRnewswire.com, and others on the “loser” list.
So what does this show? Either Google has already backed off this Panda update, which isn’t something that seems to have been noticed or announced, these sites all leaped into action as soon as they heard of this update and quickly fixed all their issues, or there was an issue with the data that week, that’s corrected itself this week.
Whichever happens to be true, the moral of this tale is that you need to make sure you take into account all evidence before jumping to conclusions and panicking. Just because you read that Panda has hit a certain site because of a couple of data points, or that Google is now favoring a particular type of site, don’t assume that to actually be the case. Don’t base a strategy on two data points that, in the long run, may mean absolutely nothing.
The press that covers SEO matters has to talk about an update when it happens, as close to when it happens, so they pretty much have to use whatever data they can find, but unfortunately, as we’ve seen here, that may not be as accurate as they’d like.
So the next time a Panda update comes out, take a deep breath, examine all the data points, and don’t panic.