After unprecedented numbers of sites recovered from Panda as a result of the last confirmed update, many people stil have questions about what sites need to do to escape the wrath of the algorithm. The answer: the same things as before, regardless of how many other sites are seeing a return to the SERPs. The 10 day update is still in full flow, and with so little data around Panda recoveries it can be difficult to know where to start.
By now, most everyone has seen a graph of what Google's Panda algorithm update can do to a site, but not a lot of people have seen what a recovery looks like. That's probably because there haven't been many documented recoveries – at least not full, 100 percent traffic returning recoveries – at least until July's update.
The good news is that partially recovering from Panda should become more straightforward, as Google reportedly moves to add Panda to the real-time algorithm, rolling it out continually for up to a third of each month, instead of on a more occasional basis. Sadly this also means that it's going to become more difficult to diagnose Panda issues – at least from the go-to resources such as Moz's Google Algorithm Change History.
A completely separate (though related) issue is the state of content not on your site. Many sites have experienced Panda problems related to content that has been scraped and hosted elsewhere on the web; as well as the more embarrassing problem of hosting content that has been stolen from elsewhere.
One voucher site my agency has been working with for a number of years was badly affected by the Panda algorithm April 11, 2011:
The site was runs an affiliate program, which at the time hosted a section on the site linking out more generally to great deals on products from across the web, in addition to the core product. These pages typically drove around 20,000 visits per month; however each deal's description was (very helpfully) provided by the manufacturer.
This meant that those small blocks of vaguely useful text could also be found in many other locations on the Internet, and there was no way our affiliate program could be the source of that information, hence the impressive drop illustrated in the graph above.
This issue was dealt with by killing this section of the site once it had been determined that this was likely to be the cause for the loss in traffic/visibility. These pages were irrelevant to the core purpose of the site, and in comparison were badly maintained.
The poor quality pages in question were redirected to another, less important site owned by the same company, and the traffic very quickly returned…and grew, well beyond the 20,000 that on paper were guaranteed losses as a result of redirecting the pages.
One of the most significant contributors to the low number of documented Panda recoveries is that while Searchmetrics is a great tool for diagnosing huge drops in visibility, it won't register massive gains in traffic once the site cleans up. You need Google Analytics for that, which means that Panda casualties are generally well-noted and Panda recoveries are not.
In this instance, many of the site's better rankings were related to the poorly converting terms on poorly constructed pages that were killed off – even though traffic improved beyond "normal", visibility is still pretty static.
One problem is that users will typically upload their car ad to as many classified sites as they can find in order to get maximum exposure and hopefully sell their car more quickly; the other is that in the UK, this vertical has a runaway market leader in AutoTrader, which gets two or three times as much inventory as its competitors.
Obviously from a seller's perspective, the former isn't a problem at all; it's the sensible thing to do. From Google's perspective we're left with an entire (extremely competitive) vertical with the same content across nearly every site. These sites are invariably crushed by Panda, and take desperate measures in order to escape, such as below:
A prominent car site recently purchased a new (dropped) domain, and redirected its previous one, in order to escape from an old Panda problem that had been plaguing the site for months.
Visually the company had created a wholly new site, with new branding to go along with it; however the same content issues lingered because the same content was hosted on the site, and the visibility plummeted once the algorithm established what had happened. Switching domains is no way to escape Panda – bad content is bad no matter where it's hosted.
The way my agency tackled the drop for the voucher site involved hemorrhaging the bad content pages from the overall strategy, despite being painfully aware of how much traffic the site could lose. The difference for many car classifieds sites is that there is little or no good content on the site to begin with, and what good content there is can be found elsewhere on the web in a more easily digestible format (on AutoTrader).
The good news is that several companies in this vertical have seen a recovery thanks to this update, as illustrated in the graph above. Many car classified sites have been rapidly improving their pages; adding useful content such as videos and reviews; and still seeing no result for many months.
The better news is that car sites naturally have access to a large, still niche audience, by leveraging their authority on several subjects: which car to buy and how to buy it. Large libraries of content sit naturally on car classified sites that many users will find helpful, and with users visiting 11 pieces of content before they convert, a large inventory of content, as well as a large inventory of cars, can be extremely helpful.
A travel site my agency had been working with for years had experienced a big loss in search traffic and visibility due to a Panda update. The site would collate holidays from other travel operators' sites, which inevitably meant that large portions of the descriptions could be found elsewhere.
Consisting of itineraries and general resort information provided by the holiday operators, nobody would suggest that this content belonged to the site to begin with – but nobody could argue that this content wasn't useful for people looking to book a holiday through the website. Rewriting the descriptions would take months, or even years, and with Panda strangling the traffic there probably wouldn't be years to work on this.
To combat the problem the stolen content was placed in iframes so visitors could still read it; set about adding more original content to the pages; and after a few months the site did recover.
The same tactic could be employed across voucher and car classifieds sites, and in a recent Webmaster video Matt Cutts said that content that is not necessarily visible to search engines is not necessarily bad; but the issue we often face in those industries is that all the content can be found elsewhere. If we put that in iframes and add more content, such as buying guides and general information, then Google is going to think our site is intended to do something completely different to what it actually does.
You can employ a content marketing agency or become a content marketer yourself, and unless you look subjectively you'll end up with a blog full of awesome frickin' content and your business won't sell itself effectively on the pages that you need people to land on and convert. The last thing you want is a blog post that convinces a visitor to buy your product – and a product page that convinces them to buy it from someone else.
Take a look at your landing pages and start asking the tough questions.
Gone are the days when you needed to hoard your link equity. If your content doesn't answer those questions, Google will know because people will probably go back and search again.
By now, most everyone has seen a graph of what Google's Panda algorithm update can do to a site, but not a lot of people have seen what a recovery looks like. That's probably because there haven't been many documented recoveries – at least not full, 100 percent traffic returning recoveries – at least until July's update.
The good news is that partially recovering from Panda should become more straightforward, as Google reportedly moves to add Panda to the real-time algorithm, rolling it out continually for up to a third of each month, instead of on a more occasional basis. Sadly this also means that it's going to become more difficult to diagnose Panda issues – at least from the go-to resources such as Moz's Google Algorithm Change History.
Is it Panda? Where to Start Looking
Internet marketers, just like users (and Google), should be able to tell at a glance whether content hosted on your site is worth reading or not. If you instinctively ignore your own images (unless it's because they're so small you have to squint to work out what they are) and your star ratings are stuck on 0, it should be time to think about improving your content, regardless of whether your site has been hit by Panda.A completely separate (though related) issue is the state of content not on your site. Many sites have experienced Panda problems related to content that has been scraped and hosted elsewhere on the web; as well as the more embarrassing problem of hosting content that has been stolen from elsewhere.
Traffic Returns to a Voucher Site
Voucher and car classifieds sites are two of the industries hit hardest by Panda. It's not unheard of that Google might take action across a particular vertical, but the idea that an algorithm might be affecting voucher sites, car classifieds sites, and the like is worrisome for people working in those industries. Scarier still for site owners is the fact that this isn't a manual action, but an algorithmic update that just doesn't like their payday loans website or others like it.One voucher site my agency has been working with for a number of years was badly affected by the Panda algorithm April 11, 2011:
The site was runs an affiliate program, which at the time hosted a section on the site linking out more generally to great deals on products from across the web, in addition to the core product. These pages typically drove around 20,000 visits per month; however each deal's description was (very helpfully) provided by the manufacturer.
This meant that those small blocks of vaguely useful text could also be found in many other locations on the Internet, and there was no way our affiliate program could be the source of that information, hence the impressive drop illustrated in the graph above.
This issue was dealt with by killing this section of the site once it had been determined that this was likely to be the cause for the loss in traffic/visibility. These pages were irrelevant to the core purpose of the site, and in comparison were badly maintained.
The poor quality pages in question were redirected to another, less important site owned by the same company, and the traffic very quickly returned…and grew, well beyond the 20,000 that on paper were guaranteed losses as a result of redirecting the pages.
One of the most significant contributors to the low number of documented Panda recoveries is that while Searchmetrics is a great tool for diagnosing huge drops in visibility, it won't register massive gains in traffic once the site cleans up. You need Google Analytics for that, which means that Panda casualties are generally well-noted and Panda recoveries are not.
In this instance, many of the site's better rankings were related to the poorly converting terms on poorly constructed pages that were killed off – even though traffic improved beyond "normal", visibility is still pretty static.
Can Car Classifieds Sites Recover?
Car classifieds sites have struggled massively since the first Panda update. The content that users are interested in, for the most part, is the selection of cars themselves.One problem is that users will typically upload their car ad to as many classified sites as they can find in order to get maximum exposure and hopefully sell their car more quickly; the other is that in the UK, this vertical has a runaway market leader in AutoTrader, which gets two or three times as much inventory as its competitors.
Obviously from a seller's perspective, the former isn't a problem at all; it's the sensible thing to do. From Google's perspective we're left with an entire (extremely competitive) vertical with the same content across nearly every site. These sites are invariably crushed by Panda, and take desperate measures in order to escape, such as below:
A prominent car site recently purchased a new (dropped) domain, and redirected its previous one, in order to escape from an old Panda problem that had been plaguing the site for months.
Visually the company had created a wholly new site, with new branding to go along with it; however the same content issues lingered because the same content was hosted on the site, and the visibility plummeted once the algorithm established what had happened. Switching domains is no way to escape Panda – bad content is bad no matter where it's hosted.
The way my agency tackled the drop for the voucher site involved hemorrhaging the bad content pages from the overall strategy, despite being painfully aware of how much traffic the site could lose. The difference for many car classifieds sites is that there is little or no good content on the site to begin with, and what good content there is can be found elsewhere on the web in a more easily digestible format (on AutoTrader).
The good news is that several companies in this vertical have seen a recovery thanks to this update, as illustrated in the graph above. Many car classified sites have been rapidly improving their pages; adding useful content such as videos and reviews; and still seeing no result for many months.
The better news is that car sites naturally have access to a large, still niche audience, by leveraging their authority on several subjects: which car to buy and how to buy it. Large libraries of content sit naturally on car classified sites that many users will find helpful, and with users visiting 11 pieces of content before they convert, a large inventory of content, as well as a large inventory of cars, can be extremely helpful.
How One Travel Site Recovered from Panda
"Cloaking" isn't an issue that often crops up on marketing blogs these days, and few SEO professionals will still try to display content to Google that isn't easily visible to users. But showing content to users that search engines can't see is sometimes necessary to escape Panda's clutches.A travel site my agency had been working with for years had experienced a big loss in search traffic and visibility due to a Panda update. The site would collate holidays from other travel operators' sites, which inevitably meant that large portions of the descriptions could be found elsewhere.
Consisting of itineraries and general resort information provided by the holiday operators, nobody would suggest that this content belonged to the site to begin with – but nobody could argue that this content wasn't useful for people looking to book a holiday through the website. Rewriting the descriptions would take months, or even years, and with Panda strangling the traffic there probably wouldn't be years to work on this.
To combat the problem the stolen content was placed in iframes so visitors could still read it; set about adding more original content to the pages; and after a few months the site did recover.
The same tactic could be employed across voucher and car classifieds sites, and in a recent Webmaster video Matt Cutts said that content that is not necessarily visible to search engines is not necessarily bad; but the issue we often face in those industries is that all the content can be found elsewhere. If we put that in iframes and add more content, such as buying guides and general information, then Google is going to think our site is intended to do something completely different to what it actually does.
Should You Still Worry About Google Panda?
Google's softening the Panda algorithm is at odds with (or perhaps as a result of) content becoming an integral part of most SEO strategies. As Panda changes, so do the tactics we employ to beat it, and we need to start thinking about content differently when it comes to commercial landing pages.You can employ a content marketing agency or become a content marketer yourself, and unless you look subjectively you'll end up with a blog full of awesome frickin' content and your business won't sell itself effectively on the pages that you need people to land on and convert. The last thing you want is a blog post that convinces a visitor to buy your product – and a product page that convinces them to buy it from someone else.
Take a look at your landing pages and start asking the tough questions.
- Is Your Content Only There for Google? Again, we're not talking about cloaking. Users can see it, but they also won't read it. Your "150 words of unique copy" might as well be white font on white background. If the intention of your landing page is to sell your product then you need to write copy that will do that for you; your content should be your sales assistant.
- Does Your Content Answer All the Questions? Find out what people are searching for – if you think people are looking for your product when they're asking a question, you need to make sure your content answers it effectively. Not many people enter "where do I buy this from?" or "should I buy this product?" into Google, and yet those are the questions SEO professionals instinctively answer in the copy: "we sell this" and "you should definitely buy it".
Summary
People are asking "why should I buy this instead of that?" and "what's in it for me?" If your content can't answer those questions, then you should link to someone who can.Gone are the days when you needed to hoard your link equity. If your content doesn't answer those questions, Google will know because people will probably go back and search again.
Article Post @ Search Engine Watch
Posts
