• http://www.spiderwritingseo.co.uk/ billmarshall

    Couldn’t agree more on pretty much everything. I regard rel canonical as a last resort and avoid it if possible.

    Said for years that attention to detail is the single most important thing in SEO (and webdev too), and agree that a holistic approach where everything is pointing in the same direction makes a big difference.

    I had a feeling that 301s weren’t as effective as they used to be but I didn’t think it was quite as bad as you say – if it is then it’s worrying for lots of perfectly valid site development reasons.

  • Duane Forrester

    Adam, Adam, Adam… oh, Adam…

    Thanks. You just made my morning. :)

    I agree with most items, though I’m not entirely sure about the depth of “value errosion” when a 301 is implmented. Perhaps in some cases, it can be this bad, but not in all, nor even, most, in my experience.

    Anyone trying to use rel=canonical as a blanket solution for dupe content needs to pay attention, too. Fix the issues with your site/feeds/cms, etc., then use rel=canonical sparingly, lest all that work you put into it become null-and-void when Google decides enough folks have dropped the rel=canonical ball and quietly start ignoring it like they have so many other directives over time.

  • http://www.audettemedia.com Adam Audette

    Bill – your mileage may vary on 301 link equity erosion. It has really been eye opening for me of late, however, so pay close attention to this. 301s may act different in different circumstances. For example, how a page that is exactly the same aside from how its URL is changed (for SEO purposes) is scored with a 301 may be different than a page actually moving for different reasons. There is a lot of subtly like this in SEO I’ve found.

    Thanks, Duane, great points as always. See above – would agree that different cases will react differently. Actually you made me start to think about this more, so thank you for that, I definitely need to do some more research and testing on this with different variables.