How do I come to this conclusion? Well, I was updating my hubpage on Google Caffeine, which I wrote on 12th August when Google first told the world about their new "next generation" engine nicknamed Caffeine.
Briefly, Caffeine is not really a change to Google's algorithm (though some slight differences appear), it's really a change in structure of the index itself. This is really important, as what is indexed impacts rankings - eg if a lot of pages that have links to your site are suddenly indexed, it should propel you up the rankings, and vice versa.
In my original analysis I was comparing pages in the current index to the Caffeine index for various topics and also for various web 2.0 domains.
What is interesting is that when I retested today, I found differences in the current index too.
Here's the bit that concerns Hubpages.
Using the site: Hubpages.com operator, here's what was returned on the 12th of August 2009:
Current Index: 1,800,000 pages
New Caffeine Index: 1,830,000 pages
Now here's what's returned on 12th September:
Current Index: 1,100,000 pages
New Index: 1,140,000 pages
That's a considerable drop. Up until I crunched the numbers, I had been pooh-poohing suggestions that Hubpages has suffered a penalty. I basically believed that if new hubs on hot money topics like credit cards were struggling to get indexed, it was down to saturation of topic, where people have mindlessly pumped out a bazillion hubs on the exact same subject, sometimes with the exact same title, (and are still doing so, if you look at the new hubs tab), and Google decided "we don't need any more".
But a 700,000 drop in what is indexed is not just about new stuff struggling to get ranked, it's about old stuff being dumped. I wouldn't have realised the sheer scale of the drop had I not originally recorded the numbers for my Caffeine analysis. Perhaps some of the pages that have been dumped are profile pages and tag pages - but that's still serious as profiles and tag pages send backlinks to hubs. I will continue to update that hub every month in order to build a data set of what exactly is going on.
As for advice to hubbers - stay clear of the "hot money" topics. They seem superficially attractive, but are a wasted effort. Better to concentrate on something obscure that is going to be indexed. Being indexed is more important than the potential CPCs of a keyword. If you can't get your page indexed, you will make zero.
P.S. Ezine has taken a hit too - see my Caffeine hub for details