LinkCounts & LinkStats From PubSub’s Only Rough Picture, So Far

There seems to be no end to blog search tools wanting you to have cool looking charts to prove your popularity. To prove the point, here’s another one.
LinkCounts from PubSub gives you an overview of top sites getting links and giving links. Want to check yourself or another
site? Use the LinkStats page. But keep this big caveat in mind that the page itself tells you:

PLEASE NOTE: LinkCounts are based on the content of a site’s feed(s). Some feeds only publish summaries (or even just headlines) that do not contain links. We are
constantly working to improve the resolution and accuracy of our published statistics. At times we may feel that our changes are significant enough to warrant reprocessing of
some or all of our historical "raw" data. These changes will most likely result in the removal and regeneration of our published historical data.

In other words, rather than counting links in actual posts, this is all apparently based only on what goes out in a feed. That’s a bad, misleading thing. For example,
here’s the SEW Blog stats for the past 30 days. According to this, we’re pretty stingy with
the links, giving out only one — ONLY ONE — over the past month.

That’s absurd. We link and link and link to all sorts of things. But since we send out 100 word summaries, only links appearing in those summaries are counted. Even then,
they have to be parsed out correctly — and PubSub’s not doing that. I know that, because it’s very common that a post will start out with a link — so the link is in the
summary. That happened many more times than once in the past 30 days. Heck, the summary of this post will have two links in it.

So, if you want some very rough idea of who links to you in the blogging world, this new tool gives you an idea, but it’s far from a complete, accurate picture. In
addition, while you see overall sites linking, you aren’t show individual backlinks. In other words, you have to hunt around in a site to find the actual link.

Postscript: PubSub got in touch to note that we don’t send links in our summaries at all — so there’s nothing there to count! But I haven’t heard back about why they mysteriously have found some to count anyway in the odd case. And the underlying point remains. Only links parsed out of a feed are counted, which will skew the stats.

Related reading

Simple Share Buttons