Content Farms Revisited: Panda Needs To Go Further

Back in February, I wrote a post about content farms  -- defining them and laying out the problem content farms pose for SEO. In the meantime, Google has released series of algorithm updates (Panda) to combat the problem. But has the Panda update really lived up to the goal of minimizing the effect of content farms in organic search results?

Demand Media Loses Traffic

So how hard has the Panda update hit content farm king Demand Media? According to an article in Forbes just yesterday, after several phased Panda rollouts, Hitwise estimates that traffic to Demand Media sites are down upwards of 40% since the beginning of the year. And indeed, in looking at the effect of Panda to reduce content farm rankings and actually promote non-content farm results, my own qualitative data across many clients suggests that the update appears to have helped non-content farm sites improve rankings.

In February's post, I gave an example of a search of "how to roast pumpkin seeds." At the time,, a Demand Media site, ranked at position three. However, a search today for the same term ranks at positon nine. Clearly a drop in organic ranking of six positions will produce a negative impact on traffic. And while this is just an example of one search phrase and one of Demand Media's many content farm sites, imagine that impact across the multitude of keywords that Demand Media's various sites were optimized for.

But... Duplicate Content Still Reigns

While I can't say that the goal of the Panda updates was to reduce duplicate content, Matt Cutts did indicate that the goal of Panda was to reduce spam. In my definition, however, spam would include multiple duplicate content sites in addition to content farms, which are essentially duplicate content on multiple sites. In other words, while Google has reduced the rankings on many content farm sites, it has not removed them completely from the index or, for that matter, the first page of results.

Take for instance the search for "how to roast pumpkin seeds" I did in February. I reported in my February post that I found 189 websites with exactly the same content as the site on roasting pumpkin seeds. Today, after the Panda updates, the number of duplicate content sites on this term is down to 125.

While Panda decreased the duplicate content for this search query by 34%, I think most people would agree that 125 sites with duplicate content is still far too many. There's more work to be done here.

The Problem As I See It for Google

The problem with duplicate content isn't a new one. And duplicate content by itself doesn't always signal to Google that the site is a content farm.

Take, for example, ecommerce sites -- notorious for simply pulling the description for a product directly from the manufacturer's site. In one example, I did a search for the manufacturer's description for the Dyson Ball DC24 and found 283 websites with the exact same manufacturer description.

But why wouldn't an ecommerce site use a manufacturer's description? It makes perfect sense to do so. If you're a retailer like Home Depot (one of the sites with duplicated manufacturer content), you've likely got thousands of products for sale on your website - who has time to edit each description? Additionally, Google Shopping results group together products with the same description (as well as other attributes) into one listing so that shoppers can directly compare products. Alter the description, and a retailer might lose its spot in the "compare prices" grouping.

And if you're Google, you could likely determine that ecommerce sites will be immune from the duplicate content rule. But how does Google know a site is an ecommerce site? There are two main ways Google could figure it out algorithmically: 1) if the site has Google Checkout installed or 2) the site is listed in Google Shopping (i.e., the retailer uploads a feed into Google Merchant Center). Both of these options might help Google better decipher algorithmically which sites are ecommerce-based, but many retailers still don't use either of those two Google services.

Bottom Line

So what's the bottom line? Content farms may have reduced rankings, but they're not dead. In fact, they're not even off the first page of Google results in some cases.

And duplicate content? It's a much bigger issue for Google to tackle, one that hasn't been handled completely yet. Content farms are a major contributor to duplicate content, but so are ecommerce sites and other types of sites that rely on similar or flat-out duplicate content for other reasons.

I do, however, have more hope now that Google will continue to fight off content farms algorithmically. I think that Aaron Wall made a good point in his post today about content farms: They made Google look stupid. And, as Aaron put it, "Don't make Google look stupid. That is the #1 rule of SEO." Well said, Aaron.

4 comments about "Content Farms Revisited: Panda Needs To Go Further".
Check to receive email when comments are posted.
  1. George Linzer from Potomac River Media, April 26, 2011 at 2:25 p.m.

    Thanks, Janet. You educated me with your first article about content farms. Now I'm wondering if a legitimate site can be negatively impacted by being referenced multiple times by articles on a content farm?

  2. Paul Camp from Content That Works, April 26, 2011 at 2:51 p.m.

    Duplicate content has one other legitimate use. Our company syndicates specialty content to websites run by local media outlets — everything from newspaper and radio station websites to independent entrepreneurs. We supply a mix of daily original content and aggregated content in 12 niche categories like bridal ( or holidays ( Ideally our local clients add original local and local ads, business directories and the like, but even if they don't localize the niche it is surrounded by other types of original local content on their sites and offers site visitors an online vertical that is as deep and rich and ever-changing as anything on the web.

  3. Janet Driscoll miller from Search Mojo, April 26, 2011 at 9:05 p.m.


    Good point... many real estate listing could be duplicate content across sites too. But I think, like ecommerce, Google will likely need to come up with a way to identify these sites. It will be interesting to see just how far Google will go to reduce duplicate content and content farms. Like I said, this is a good start, but there's so much more that could be done.


  4. Janet Driscoll miller from Search Mojo, April 26, 2011 at 9:09 p.m.


    I think that is the great fear and possible reality... Google has to be very careful when treating things strictly algorithmically, because how can an algorithm always ensure that sites are one type or another? It's a big challenge.

    The way legitimate sites can avoid it is simple... use the canonical tag if you have duplicate content WITHIN your own site. If you have duplicate content across multiple domains, I think you do risk appearing as a content farm. But I also wonder what other factors Google looks at to identify a content farm? Something to keep an eye out for as Google continues to try to reduce content farms through algorithm updates.


Next story loading loading..