The PR Friendly Index has been good to me. I initially compiled it as an ongoing experiment to see how I could ‘measure’ blogs, especially en masse, especially using forms of automation that would make it as easy as possible. The ultimate goal was something along the lines of the Power150, except I had visions of creating a blog ‘index’ akin to the FTSE-100, so we could see who was rising or falling in real-time.
This never came to be, not least because I had neither the time nor the expertise to make it so. But I got a fair way with canny combinations of Google Docs XML and ImportHTML calls, Technorati APIs, and good old-fashioned Word macros to format everything.
Since then, of course, we all know that this isn’t the right way to ‘measure’ a blog. It’s more qualitative than quantitative. But I still thought there was value in compiling it from time to time, provided I could do this regularly using proven processes and techniques. Not least because, as I said, it’s been good to me. My stats shot through the roof when I started it, and it’s helped me professionally in many ways.
However, the last PR Friendly index I compiled, in March 2009, was a right old pain in the arse. It seemed that the Google Docs calls weren’t reliable, even though they were doing things ‘properly’ using my Technorati API key. As a result I had to copy and paste many results manually, or estimate some results much in the same way scientists inserted frog DNA into dinosaurs in Jurassic Park.
I’ve recently been toying with the idea of starting the index again, and took a quick look to see whether things were still intact. The bad news is that they most decidedly are not.
Take a look at the image below. It’s the result of the Technorati calls for the authority of the first few blogs in the PR Friendly Index:
See that? Most of them don’t work. There’s nothing wrong with the calls. The blogs exist. If I look them up using the same API calls as in the Google Docs, but typed directly into the browser address bar, they work. Do the same through Google Docs, and they don’t. Mostly.
So I can only imagine there’s something going on with Google Docs calls. I get similar results when using ImportHTML to query Google Blog Searches. Hardly any of them work.
If none of them worked, I could start debugging this. But some of them, working some of the time? Nothing worse for debugging.
This leaves me with a problem. If I want to continue with the PR Friendly Index, I need to figure out yet another workaround because I simply cannot spend the best part of a day compiling these figures manually. Something’s gone wrong, and I need a fix, ideally a nice efficient way using API calls.
Can anyone help?