PR8 Site Lost 50% Traffic This Week

[alika]

A PR8 site I consult on lost a little more than 50% of its traffic from Google this week. I don't know if this is part of the Panda update but this site is not a content farm

All original content, doesn't use anybody else's content. No ads, except for their own books and products. Main site hardly links to outside sites, except as footnotes and citations. Has several content products - ranging in words from 1,200 to 25,000+ in depth studies.

They suddenly disappeared in many of their top keywords. They don't even show up anywhere for their experts' names, which they were always at the top. Google crawling has slowed down in March from a high of 30,000 in Feb down to 6,000.

One thing to note is that their experts contribute op-eds and regularly published in WSJ, NYT, Washington Post, Washington Times, etc. Their deal with newspapers is that these publications have the first dib at publishing the op-eds, and they publish the op-eds on their site 24 hours later. Of course, when WSJ for example publishes it, other blogs/aggregators follow suit and by the time this site publishes their own content, it doesn't show up anymore in Google

Their blog, which is a subdomain, doesn't seem affected. It is in top 100 of Technorati politics list.

The site made a massive redesign March 2010 and changed their URLs. Traffic has dropped after the redesign, and a year later still hasn't recovered.

But this week's decrease is different, sudden and massive.

What are the things I need to start looking at?

Thanks guys

[marketingmagic]

I'd suggest you submit the site for review in the [webmasterworld.com...] section.

With so many potential variables one would have to actually look at the site to provide you with any insights.

That being said - the "shared" articles raises a red flag in my mind without even looking at the site.

[goodroi]

A PR8 site I consult on

Are you an SEO consultant or another type of consulant?

You really are not providing enough information to get back good feedback.

-Have the backlinks changed?
-Did they do any link development?
-Do they sell/buy links?
-Have you checked to see if the traffic loss is due only to a drop in Google referrals?
-Have you scanned for malware or signs of hacking?
-Have you checked to see if another site stole content?
-Are there any 302 or 301 issues?
-Have you checked for messages in WMT?

I could go on and probably list over 100 possible issues. If you want a better response please do a little more research and then ask for insights. It sounds like you have many issues if you have been consistently losing traffic since the 2010 redesign. Good luck.

[martinibuster]

I just want to make sure I understand this. Is the following correct?

• The PR 8 site creates content.
  
• The content is first published on authoritative news media sites.
• Then it's published on the PR 8 site.

[TheMadScientist]

martinibuster, you're not suggesting the order of the posting of the content makes this site about as close to the text-book definition of a content farm as I can think of, are you? I mean it's their content, they're just the last one's to publish it...

Sorry for the tongue-in-cheek bit, alika, but it seems pretty obvious where the issue is if they really wait 24 hours to publish content others such as the NYT, WSJ, etc. are publishing and having scraped in a much more 'timely' fashion ... They're way late to the news game, even if it's their content.

Out of curiosity, what do they expect?

Really, they re-publish 'news' content 24 hours after it's news and they expect a search engine to consider them the source of the content and rank them for it? I'm all confused¿?¿

[edited by: TheMadScientist at 8:22 pm (utc) on Mar 25, 2011]

[econman]

The licensed content/duplication problem you describe (the content is initially published as an op/ed on a major/authoritative site, then scraped/duplicated by others, then finally published on their own site after the embargo period) certainly sounds problematic.

What % of the pages on the site fit this pattern? (Not including the blog, which you say hasn't seen a sudden traffic drop).

[Simsi]

Main site hardly links to outside sites...

Might that also be problematic? I would have thought that fitted into a footprint of a content farm also. Some authoratative and relevant accompanying links would be quite usual for a site like this wouldn't it?

[tedster]

If any site manages to get to PR8 they were doing a lot of things right. I notice that you said the loss came this week - so that doesn't sound like the Panda Update to me. I'd lean first toward either cloaked, parasite content being hosted or a very nasty competitive attack. Maybe someone on staff has been selling links on the side, too - hey, it happened to Google once.

I'd also drill down into which keywords and landing pages took the biggest hit. You need to find a loose thread and just start unraveling it. And yes, the Site Review forum in our Supporters Area [webmasterworld.com] may be the best place to get real, actionable help.

[alika]

Are you an SEO consultant or another type of consulant?

More of web analytics

-Have the backlinks changed?

Backlinks have dropped significantly after the redesign and they changed the URL. They did not go back to ask other sites to link to the new URL. They also did not do a very good job in redirects (e.g. many of their internal pages were redirected to the homepage, not to the new URL of the page).

-Did they do any link development?

No. The don't do any links. Just organically. They're more focused on traditional marketing - with a huge communications team divided into TV, radio and print. They now have social media, but not really SEO.

-Do they sell/buy links?

No. They do buy advertising spots (e.g. ads in the DC Metros, banner ads in Drudge, etc. but not links).

-Have you checked to see if the traffic loss is due only to a drop in Google referrals?

Yes. Only Google. Overall traffic dropped only 18%, as there are other sources of traffic. Google dropped 50%

-Have you scanned for malware or signs of hacking?

None

-Have you checked to see if another site stole content?

Yes. Many. And they have never been one to file DMCAs at all.

-Are there any 302 or 301 issues?

One issue I've seen is that for some searches, an internal URL that they use to view any changes of their website (e.g. instead of www.URL.com the internal URL is origin.URL.com) is now visible in Google. Instead of the www.URL.com version of the page, it is the origin.URL.com of the page that is showing up. The www. URL is nowhere to be found

-Have you checked for messages in WMT?

Yes. None

[martinibuster]

Is the following correct?

• The PR 8 site creates content.
  
• The content is first published on authoritative news media sites.
  
• Then it's published on the PR 8 site.

[edited by: martinibuster at 9:07 pm (utc) on Mar 25, 2011]

[alika]

Sorry for the tongue-in-cheek bit, alika, but it seems pretty obvious where the issue is if they really wait 24 hours to publish content others such as the NYT, WSJ, etc. are publishing and having scraped in a much more 'timely' fashion ... They're way late to the news game, even if it's their content.

I know. I've told them a lot of times. And the traffic of those op-eds are really really pathetic on their site.

But it's a deal they've worked with the newspapers back in the 80s and 90s. They'd rather increase their chances for pickups of the top print publications and have WSJ, WP, Scripps Howard, and the likes use their op-eds than make sure that they are the first one to publish these op-eds in the Web. And they've been very successful in getting print pickups

I've been concerned about that. Because when WSJ publishes something, the same piece is regurgitated by other blogs/sites/farms

I've already talked to the new VP, and he is concerned with the very low traffic of these op-eds. So they may be more open now to changing their procedure compared to the old VP.

[alika]

1. The PR 8 site creates content.

Yes. They produce books, in-depth studies, quantitative/mathematical papers, shorter pieces, and op-eds. Only the op-eds are shared with other sites.

2. The content is first published on authoritative news media sites. Then it's published on the PR 8 site.

Only the op-eds. Not the other types of papers. In a month, they'd produce about 100-300 pieces of content. Op-eds are about 10% of the content produced per month

[TheMadScientist]

Why not noindex the op-eds that are published by the papers? Then at least you don't look like you're trying to rank for 'yesterdays news' today...

[alika]

Thanks MadScientist. I'll certainly raise that with them

Given that this is a well established name in its sector, Google is only 40% of their traffic. But doing the site operator shows that Google only has 12,000+ of their pages, compared to Yahoo which has 70,000+. They don't have a working sitemap in GWT since the redesign a year ago -- their IT just didn't think it's a top priority (I know, horrors!)

Some patterns re their traffic drop:

1. Traffic for branded keywords remain strong.
2. Keywords where the landing page is their home page are not affected
3. Keywords for their key products (books) are not affected
4. Keywords most affected pertain to the old papers (2005 and older). These are the oldies but goodies that get the most traffic
5. Names of their experts (those with long bios + long list of papers published) have disappeared; other staff member pages including former experts are still ranking at the top

Traffic for some keywords started falling on Sunday, but most fell on Tuesday.

[incrediBILL]

See this thread [webmasterworld.com] as well, the OP thinks that dated content may be a target and I'm seeing similar issues.

[alika]

One other problem with the site is that they use Amazon Web Services to generate PDF versions of their papers. Now when searching for exact titles of their papers, the site's pages no longer comes up. BUT the AWS URL of the PDF of the paper shows up in the SERPs

They started this practice a year ago after the redesign. Could this be a duplicate content penalty?

[crobb305]

the OP thinks that dated content may be a target

Can you elaborate on what you mean by "dated"? Do you mean aged (old) content, or articles that actually have a date stamp on them? If you mean the latter, I wonder if a javascript that displays today's date in the template could get cached in older content and cause problems.

[tedster]

Here's a thought - a possible way to address the fact that the syndication sources are outranking your site. I assume, since you are being syndicated by major news outlets, that your site is also a Google News publisher. Have you considered using the experimental "syndication-source" meta tag for News sites [googlenewsblog.blogspot.com]?

[alika]

In our case, the drop off may have started with older papers. But now, even the newest papers published as recently as this week are no longer appearing in the SERPs.

It used to be very quick for Google to pick up new papers for this site. Now, searching for the exact title of the papers are not yielding the page. Except for:

- the PDF version in the Amazon Web Services URL
- For papers that are still in the homepage's list of most recent papers, then the homepage of the site shows up
- the RSS feed of the issue area, if it still has the paper
- all the other sites that copied the paper

[incrediBILL]

I had something similar happen when I added an RSS feed that Google bizarrely started attributing all my new content to the feed instead of the pages listed in the feed.

Had to block the RSS feed in robots.txt and let Google get updates from FeedBurner instead and everything was cool.

[alika]

Thanks Tedster. Yes, the site is in Google News. Interesting that a paper published last Friday could be found in Google News, but not in main Google.

I'll definitely suggest that metatag to them.

Incredibill, this is actually the first time I am seeing their RSS feed ranking instead. And no, their feed is not using Feedburner.

For a big site and organization, their web operations is still a mess. Heck, they are using Addthis -- and no one has seen the stats of their Addthis because no one seems to have the login information to the account.

[TheMadScientist]

...and no one has seen the stats of their Addthis because no one seems to have the login information to the account.

That was so good for a laugh...

[tedster]

For a big site and organization, their web operations is still a mess.

That's so true for many big organizations - especially if they are not pure-play web but have a lot of offline activity. If the offline part of the organization pre-dates the web, the org chart and internal communications rarely are agile enough to cope with the online needs.

[ken_b]

and no one has seen the stats of their Addthis because no one seems to have the login information to the account.

What about the weekly summary that is emailed?

[alika]

What about the weekly summary that is emailed?

I've searched high and low and no one admitted to getting the weekly emails. Oh, it's the same thing for their Feedburner account, iTunes account for the podcasts, and others.

The organization's setup focuses heavily on offline marketing -- e.g. 8 person team for marketing to TV stations yet internet marketing dept existed for only 1 year before being abolished and just combined with the social media team where the main focus is social media.

[alika]

One culprit could be wrong canonicals. Last Thursday they made a change in their CMS and uploaded a wrong canonical URL for all pages.

They're going to fix that today. Hopefully it will solve some of the problems

I'm not totally sure if the canonical really caused the drop, or just the final nail in the coffin, because:

- the wrong canonical was placed throughout the whole site, yet some pages did not disappear (e.g. homepage, pages of non-experts which have less authority than their experts pages)
- the wrong canonical was placed Thurs. but keywords started disappearing Tuesday.