• http://www.seoskeptic.com/ Aaron Bradley

    A perfect storm of hash bang URLs and canonicalization issues. And let’s not forget Twitter’s handling of secure pages (or lack thereof). If I conduct a site:site:twitter.com/vanessafox query on Google I actually get the following snippet:

    Vanessa Fox (vanessafox) on Twitter
    https://twitter.com/VanessaFox – Cached
    Vanessa Fox (vanessafox) is on Twitter. Sign up for Twitter to follow Vanessa Fox (vanessafox) and get their latest updates.

    Note the https:// – egads!

    Now that Twitter resolves to the hash bang form of the URL for human visitors, I wonder too at the SEO impact of all those hash bang URLs encoded in links. As you say, if you’re going to go to all that trouble to resolve to the #! version of URLs for visitors, why bother with the redirect to the non-#! version for search engines in the first place?

    The other curious (baffling?) thing about how Twitter presents its pages to search engines is how the bio is presented to the search engines. The and are both fairly useless templates (full name added to the H1, user handle added to the ). The really potentially important piece of static information about the user (the bio) is relegated to a list in the code, and it isn’t structured to be strongly associated with the user’s full name or Twitter handle. A phrase search for almost anyone’s bio content (e.g. “fascinated by our evolving online searching culture”) invariably brings up sites (or Twitter lists) referencing the bio content, rather than the relevant twitter account itself.

  • http://www.seocharlie.com/blog Carlos Chacón

    Great analysis Vanessa! No doubt Twitter is having some issues that collaborates to don´t get search engine visibility. Ironic because it has tons of content of each account that can show in many different ways to the SE.

    I just wonder if they really know about all of that!

  • http://mikecanex.wordpress.com/ Mike Cane

    It’s interesting you mention the rate limit.

    I recently used TweetScan. It did the best job of any service of grabbing my tweets, supposedly going back to day one (but not tweet one).

    However, yesterday I opened my backup file to search for a tweet I *know* I sent out — and it wasn’t there.

    So either:

    1) Twitter has lost tweets or
    2) This rate limit also affects everyone’s ability to get a backup of all of their tweets

    TweetScan mentions up front that rate limit could cause some backups to take a while. But now I also wonder if it affects their — and all other services — ability to really get all tweets out.

    Or maybe the issue you bring up here could account for that.