Welcome to the Q&A Forum

I will start out by saying I am not answering your question, really. The whole reason I found your question here is because I was about to jump on and ask THE VERY SAME QUESTION. I am working on several projects, and on at least 75% of the projects, I am seeing the exact same thing. The one or two sites I am targeting in OSE are CRUSHING ME with what Google and Cutts would perceive as crappy backlink sites. I have seen some people have responded to your question by saying "don't let OSE, Raven, Majestic or whatever" be your guiding light in this regard, as Google doesn't use any of these services. But the pattern cannot be denied. Backlinks STILL work. Now, it's possible this could change at any time, and I am - for the most part - refraining from duplicating their (in Google's opinion) backlink strategies on these thin directory sites. However, it is really difficult to explain to your client why slow and steady will be better in the long term for their SERP efforts than blasting their link and anchor text out to dozens or hundreds of sites. Of course, there is this - which was pointed out on inbound.org I believe: https://support.google.com/webmasters/answer/34432?hl=en Before May 27: In general, webmasters can improve the rank of their sites by increasing the number of high-quality sites that link to their pages. After May 27: In general, webmasters can improve the rank of their sites by creating high-quality sites that users will want to use and share. == With this in mind, I think a lot of those sh*tty backlinks that are beating my arse are likely on their last leg. Again, this is all conjecture. But that sly change with no real announcement is interesting, no? edited for formatting

Keyword research isn't the primary driver.  But I check to see if the queries have adequate volume to justify the time required to produce the content.

Like Brad said, just try to retain the URL structures as much as possible unless you are really aiming to make it more seo friendly. Do the migration quickly and make sure to minimise broken pages. It'll usually affect your rankings, but you'll recover. Just do cleanup quickly and have your linking plan ready

Hello! Phil just flagged this up to me. Great Question. The keyword level targeting for video is no where near as specific as Search advertising. I would say even Display Network keywords have more accuracy (which is saying quite a bit about how horrible the YouTube keyword lists are). Phil's suggestion is excellent. I would do that in addition to taking these steps: I would first start with a retargeting campaign with video. You can make the list in Analytics and bring it into AdWords. Then I would make a target group based on Topic & Interest (topic being the content of the video, and interest being interest of the user). You can always go in and see where your ads are appearing and exclude, like the Display Network. But, the first half of your question was referring to ROI and tracking. It's tricky. You have to override the system. Here's how: Step 1. Create a mirrored landing page...a version of an existing page which you can direct all video traffic to with an unique URL. eg www.DOMAIN.com/NewLandingPageStep 2. Tweak the GA code (ONLY on the mirrored landing page) to something like this http://www.youtube.com?utm_source=google&utm_medium=cpc&utm_campaign=VideoAdsFull example below var _gaq = _gaq || []; _gaq.push(['_setAccount', 'UA-123456789-1']); _gaq.push(['_setDomainName', 'Domain.com']); _gaq.push(['_setReferrerOverride','http://www.youtube.com?utm_source=google&utm_medium=cpc&utm_campaign=Video']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); (no matter how hard I try, this won't format the way I want it to. I'm sorry that looks so shitty) Regular video traffic from the GDN should be tracked normally but if not then creating a second page and having a different dummy referrer should be easy enough. This is the hack that one of our Google Reps provided us. I don't have much information on ROI. I haven't dedicated a great amount of time to it (which is something I need to do and figure out this month). Hope this has helped!

In December 2008 I helped a company transition from BV4 to Volusion. Like your situation, our URLs were years old at that point (the original Website launched in 1996). I really only have 2 pieces of advice: 1. Make sure you've set up your 301 redirects properly (we didn't, and it took 6 weeks to sort out the mess). 2. Hang on for the ride. Be prepared for a major dip in traffic, particularly if you are re-writing all of the URLs. But take heart. Sometimes the dip is necessary to move onward and upward. Seth Godin wrote a whole book on The Dip. Now might be a good time to buy your CEO a copy. Cheers! I hope all goes well and that it moves you one step closer to awesomeness! Dana

Thanks Robert. Yes, I believe we have a good link profile and work hard at it. This is really a shame. I don't see any reason to really waste more time and not submit a rec request to Google in the next few hours. I will keep everyone posted and thanks for contributing.

Ha! I've seen those far too often. I don't know the process that led a company to get a contract by building links like that. Frankly I'd be interested to see their proposal. I was given the opportunity to look at competitor proposals on several occasions. Most bank on giving an informational overload to show that they have top-secret SEO tools and knowledge of the industry. I saw a 40-page PDF spitting out data from what seemed like a normal SEOmoz account. I think they get the deal-maker to focus on data and "wow" them instead of putting focus on case studies and guarantees. I just can't imagine why a proposal would ever be over five pages. I think Alan Weiss also states that all proposals should be under three pages. Harder to do sometimes given large projects, but still a good rule to follow. Cheers to tasteful back linking.

Ooo, Panda, that's caused a fair bit of trouble hasn't it!? We did the following; Re-formatted and re-wrote a lot of old or 'rubbish' content. Including it's layout and the navigation process Culled any links which looked a bit spammy and tried to obtain links from site with higher MozRank and more credible links itself Updated all our business listings on the major listing sites such as Yell, Google Local, etc. And made sure our domain and addresses on the website corresponded with the business listings. With a view to increase the overall trust of the site. Hope that helps!

I pulled out some metrics on a few http://www.garethjames.net/authority-directory-list/

Agree with EGOL.  if you can create some amazing linkbait content - think viral, creative, or awesome infographic you could in theory leverage this to a short burts of traffic, inbound links etc.  depending on the site and content you could leverage on digg, reddit, stumpleupon, etc. the problem with these social sites is that you typically have to be an active member to start leveraging links and information for real value.

Participate participate participate.  If you are sincere and leave great comments it will help.

I have a lot of thoughts on this subject. If I was to make a blog entry on this topic, it would span multiple pages or have to be broken down into sub-topics. I do think there is a correlation between good code and search engine rankings. I do not think there is a correlation between a w3c validated page and search engine rankings. The validator is not current enough, nor flexible enough, to accommodate the real world situations which websites encounter. Example A: HTML 5 is recognized by all major browsers. W3C validation of HTML 5 is still experimental. A specific example that applies to SEO is the canonical tag. According the the W3C validation site, the canonical tag is not currently valid. Take a snippet of HTML5 code which passes validation, add a canonical meta tag to the header, and the code will no longer pass validation. This is a direct conflict between best practices and validation. Example B:  The world's most popular web page, google.com, does not pass validation. Matt Cutts discussed the topic. In short, they had a choice between providing code which validated, or code which worked. They chose to go with the working code. Example The standard facebook widget code, youtube video code, and other popular code does not pass validation. Whenever I design a website, I check the code in the validator to look for errors. Initially, I will find numerous errors related to code outside of my control such as social sharing widgets or youtube videos. Once I remove that code, the page often validates. I have researched the issue and it is possible to modify the facebook code or youtube code so that it still functions and passes validation. Doing so requires extra effort, it provides absolutely zero benefit other then saying "hey, I pass validation", and there are often drawbacks such as having to add extra javascript to your site which can otherwise be viewed as unnecessary code.

Hi Ryan. Glad to hear someone talking sense! In researching sites' linking patterns I have also often come across sites that are linked to that should really have no logical relationship back to the site in question, yet the link still counts. Sometimes these are from blog sites that seo companies have set up themselves, that have little content, and not terribly interesting content either, yet this still seems to work in 2011, which is baffling. I agree that one day this will all be devalued, but you wonder when.