What the Panda 2.2 Update Means

It’s been nearly four months since Google’s Panda update first made its presence felt. Dubbed “Farmer” by some, it hit content farms particularly hard. With Google rolling out the third update to Panda, what do we know about this bear that can keep you from getting clobbered?

Barry Schwartz at Search Engine Land reported on the new version of Panda. He noted that Matt Cutts reported at the SMX Advanced Conference that it had been approved, but not yet rolled out, and that he started seeing hints of its influence on SEO forums on June 16. On June 20, he received confirmation from Google that the ranking changes being noted in the threads were probably due to the Panda 2.2 update.

A lot of sites that weren’t previously affected by Panda are feeling its wrath now; many others that suffered under earlier Panda updates are actually doing a lot better after the change. Whether or not you’ve been adversely affected, these changes can be unsettling. Are we back to the days of the Google Dance?

It may look like it at first, but we’re not even close. According to Danny Sullivan, what we’re seeing is nothing like the old days, when Google would update its index every month or so by “suddenly dump[ing] millions of new pages it had found into its existing collection” which “caused ranking changes that could take days to settle down, hence the nickname ‘Google Dance.’” We’re also not seeing a full-fledged update to the algorithm, along the lines of “Florida” in 2003 and “Mayday” of last year.

So what exactly is going on? To understand that, you need to understand the nature of Panda. The nearest comparison Sullivan could make was to PageRank. Yes, PageRank – that factor that supposedly has no effect on your ranking in the search engine results pages and which Google rather wishes we’d all ignore. Panda, he says, is not itself an algorithm change; it’s a ranking factor.

Before you start scratching your head, let’s review how search engines work and define a few terms. Search engines send out software to examine web pages for indexing. They look at various factors on the page to determine a page’s relevance for any particular search; Google looks at anywhere from 50 to 200 of these factors, depending on who you believe. They may include such information as what keywords are in the page’s title, how many links point to that page, what words are in the anchor text used by the links, and so on.

Google’s algorithm looks at these factors. Various factors receive different “weights” or levels of importance. For instance, the meta keyword tag isn’t considered important at all, but several factors pertaining to links receive a lot of weight when the algorithm does its thing. Once the algorithm weighs the factors for a particular page as related to a search using a particular keyword, it decides where to put that web page in the SERPs.

So what does this have to do with Panda? Panda is yet another ranking factor. I can’t prove it, but I think what’s going on is that Google looks at a number of signals, not just one, that its engineers believe indicate the “thinness” of a web page’s content. I believe that Panda takes these signals as a whole and applies a formula to return a value. If the page stays below a certain value, Panda leaves it alone; go above that value, and Panda lays some kung fu on it.

Some may wonder how Panda can be just a ranking factor rather than an algorithm change when it seems to have such a disproportionate effect. I’m with Sullivan on this, though. Consider how much of a difference links make to your standing in the SERPs, and you’ll see why.

I already explained that ranking factors can have various “weights” in Google’s algorithm. They can be of no importance, like the meta keyword tag; or very important, like links from authority sites; or somewhere in between, like relevant title tags. Right now, Panda is dialed way up in importance, which is why it’s having such a huge effect. That will continue to be true for quite some time, so you might as well get used to it.

There’s one other thing which Panda and PageRank have in common: they’re updated periodically. What I think is going on is that certain ranking factors are more complicated than others, because a number of different pieces of data go into them to return a value. When enough pieces of data go into calculating a particular ranking factor, it takes too long to find those values fresh every time. So Google calculates those values for a page and stores them, and uses the stored values every time it runs its algorithm…until it performs an update to calculate new values for that particular ranking factor.

These updates happen when someone at Google decides they need to happen. Sullivan notes that they’re sometimes called “manual” updates, but that’s really a misnomer. What really happens is that someone decides it’s time to run a particular computer program to update the values. That program might not have changed since the last time it was run, but pages on the Internet certainly have, which means that the ranking factor for the pages will change – and so will their standing in the SERPs.

What I think is going on at Google isn’t just that Panda is being run again at regular intervals; I’m certain Google itself is examining the results it gets every time it runs Panda, and tweaking the Panda program based on how close they come to what they want to see. If it helps, think of Panda as an algorithm within an algorithm.

Because Panda is run periodically, if you make changes to your website based on how badly you got burned by the last update, you probably won’t see significant improvement in your standing in the SERPs until the next update. And trust me, you can expect to keep seeing regular updates for a while. Panda 2.0 got pushed out seven weeks after 1.0. Panda 2.1 popped up four weeks after that, and Panda 2.2 showed up five weeks after 2.1. Sullivan seems to think that Google will continue running Panda updates on a four-to-six-week schedule.

But don’t focus everything you do on your website around Panda updates. Sullivan reminds us that “Google makes small algorithm changes all the time, which can cause sites to fall (and rise) in rankings independently of Panda.” Even if Google doesn’t change the algorithm itself, but updates a ranking factor that feeds into it, your website’s ranking might change.

So what should you do if you see your rankings drop? First of all, don’t freak out. Second, note the date. If it turns out that the date your rankings dropped coincided with a Panda update, you might well have been mauled by Google’s newest beast. Before you do anything else, however, Sullivan suggests that you “Look at the overall traffic that Google has sent you.” You might even find that your numbers have gone up. Remember that rankings are a means to an end, and you might still be achieving that end after all, even if you’ve lost what you thought was an important ranking. Good luck!

Google+ Comments

Google+ Comments