daveambrose at data.insights.ideas brought this to my attention. I would have reblogged directly but I find that Quote posts get a little funky to decipher once there’s a good amount of added commentary.
Melissa Chang’s Why I’m kissing Tumblr a sad, sad good-bye is already a quick read (so go read it), but here’s the gist:
I didn’t know this about Tumblr. I didn’t know that the pages wouldn’t be indexed well (or show up high) on Google. I knew that Tumblr doesn’t have comments. And I knew that Tumblr didn’t have a search engine built in. These things I decided to live with.
But I didn’t know that Tumblr had a search engine optimization (SEO) problem.
I’m not an expert on SEO for a variety of reasons, but I wouldn’t jump to these conclusions at all. My thinking is that any problems are expressly related to how one’s template is designed and content is constructed — specifically in terms of relevant meta-tags, page titles, and usage of key words within text posts.
Melissa compares her Google referrals from a Tumblr blog to a WordPress blog, and this comparison actually helps illustrate important differences between them. One of the relevant WordPress blogs has meta “description” and “keyword” elements in the html, while the Tumblr blog in question does not. Again, I’m not an expert on SEO, but I assume these are basic starting points.
I suppose we could say that some basic SEO-related tags are more standard in WP theme output than they are in Tumblr themes, but there’s nothing stopping anybody from rooting around in your tumblelog theme and getting the exact same output.
Finally, there are some stats on Google referrals to the Tumblr blog vs. the WordPress one. Check out the post for specifics, but here’s one nugget of data: From March 13 to May 8, her Tumblr-based blog got 17 Google referrals. That is pretty sad, and I can’t begin to explain it.
For the exact same period of time, Google sent topherchris.com 772 referrals with 385 distinct queries. I’m not going to rest my case on this statistic alone, but I do think it’s constructive to look at a wide range of user experiences before any sweeping proclamations are made.
Comments? Am I missing something?