It’s been more than 5 months since my open letter to Google lamenting the state of their search results and how web spam was getting the best of them. And it’s been a mere 3 months since I wrote about how much Google (and Bing) loved anchor text spam.
Since that time, a number of very public search result adjustments have been made by Google. JCPenney (one of the companies I referred to in my February link spam article) was publicly outed by the New York Times and subsequently penalized by Google. It took Bing a bit longer to respond, but the JCPenney website seems to be missing from there now as well.
Also in February, Overstock.com was outed by the Wall Street Journal for providing discounts for .edu websites if they would add keyword-rich anchor text links to Overstock.com. Their penalty seemed less severe than the JCP one as I continued to see Overstock.com pages show up for numerous competitive keyword queries at Google. Recently they claimed that they removed the offending links and their penalty has been lifted (it’s not clear how they know this for sure).
And then came the much talked about Farmer / Panda Google Update, which may or may not have been in response to all the negative press that descended upon them. While many sites lost a large percentage of their Google traffic due to the update, it has certainly not stopped all web spam from working.
A quick look at the Google search results for the search query I used in the Anchor Text Spam article shows most of the same sites (besides JCPenney) showing in the results — many of which still have lots of questionable links pointing to them. The good news is that the spammiest of them are on Page 2 or later in Google’s results. While it appears that link spamming can still work, it may not be quite as effective as it used to be — and that’s a good thing.
Web spam can take many forms, though, with anchor text spam being just one of them. Others include exact match keyword doorway domains, madlib spam pages, and keyword-stuffed content (either hidden, partially hidden, or in plain sight).
Today I decided to see if those forms of web spam can still bring high rankings to sites post-Farmer / Panda. I had a ready supply of spammy sites to look at thanks to many of you filling out my web spam report form. (Keep ’em coming!)
Here’s what I found:
Exact Match Keyword Doorway Domains
This seems to be one of the easiest ways to get decent rankings these days, which to me is crazy — it’s 2011! One of the sites reported to me had teeny-tiny keyword-rich links in the footer of their main site that pointed to their exact match keyword doorway domains. The doorways looked as if they were straight out of the old Web Position Gold doorway page generator program from the ’90s! I was both amused and appalled to see such a thing still working. I know that Matt Cutts from Google has said they’re looking into the exact match domain problem, but I have to wonder when they’ll actually pull the switch, or if they simply don’t consider the technique to be spammy. To be fair, this technique seemed to work best for the less competitive phrases, which I suppose is some progress.
My Recommendatíon: In light of what I’ve seen with this type of web spam, I’d have to say that if you don’t mind that it could be a short-lived traffic generator, it’s probably a great technique if you’re starting a new website, especially if it’s going to be an affiliate site or drop-ship type of site. I’m even considering trying it for myself if I ever get the time. My hope is that Google will close the loophole before I ever get started, but it’s been working great for so long now, it’s hard to imagine they’ll be killing it any time soon. [Side note: Yes, you just heard me recommend a spammy SEO technique…hey…if you can’t beat ’em! That said, I wouldn’t ever recommend this tactic to my clients or anyone with a “real” company.]
Madlib Spam Pages
I talked about this a bit in last week’s article on getting found locally. A number of the spammy sites I looked at were using this technique. To review, “madlib spam” is when you have the same exact content, but just switch out a certain keyword phrase to create hundreds of nearly duplicate pages. For instance, many do this to target different geographical areas by simply substituting the city and/or state name. And others may substitute a certain make or model of product.
In many ways, these sorts of pages make sense. After all, most people will only see the page that has been targeted to what they’re looking for and not all the other dupes — which is most likely why it often works. Matt Cutts has spoken out against it for many years, however, which leads me to believe that Google is simply looking the other way on these. That said, they aren’t as effective for local search now that Places Pages are featured so prominently in Google’s results. But that doesn’t stop the make/model madlib spam from working.
My Recommendatíon: As with the exact match domains, I would only suggest trying this if you understand that it could eventually come crashing down on you. I’d also suggest that if you do it, you pair it with Google AdWords so that the pages serve as paid search landing pages. That way, you can at least play dumb if you get caught. 😉
A few weeks ago, a client pointed out to me a business that “must really know their SEO.” Turns out it was an all-Flash site that had hidden keyword-stuffed content. I was quite amazed that this old trick still worked and felt compelled to report it to Matt. To me, hidden content and keyword stuffing are the worst sorts of spam. Therefore, my suggestion will always be to don’t ever, ever, ever, ever do it!
As of this writing, I haven’t yet reported any of the Rat Out Your Competitor sites to Matt, but when I checked today for the keyword-stuffed one I had reported, it was completely gone from Google! I’m assuming that Matt agreed with me on that one and took action against it. While I appreciate the fact that he will do something on a case-by-case basis if I personally submit something, that’s not going to make a dent in cleaning out most of the web spam.
I should also acknowledge that dealing with web spam has got to be a monumental task for Google. In reviewing the supposedly spammy sites that were submitted to me, many were not all that spammy. I’d say maybe only 20% were ranking highly in Google due to their shady SEO techniques. In many cases, there simply weren’t a lot of better sites to show instead, which meant that the reported sites were likely ranking despite their spam, not because of it.
Unfortunately, spammy SEO techniques are so common today that it’s hard to find decent sites in competitive industries that aren’t doing it to a certain extent. Until, or unless Google catches more of it on their own, spammy SEO is going to be a sad fact of life.