Jon Udell took a look at the subscription growth in techblogs and sees a generally decelerating growth curve.

In fact these numbers are rather hard to interpret. What Jon was looking at is the number of subscribers recorded in bloglines.com (an RSS feed aggregator). What Jon is wondering about is the number of actual readers, which is not the same.

I have been trying to puzzle out the other side of the question. How many people actually read my weblog?

Looking at the server logs I see crawlers from search engines. Some readers come in from a query to a search engine, and are easy to spot. Some of the search engines can present cached results from the last crawl, and I have no way of telling how many readers are looking just at the cached results.

Looking at the server log I can see aggregators (like bloglines) and desktop readers sucking on the RSS feed. I am not familiar with all the products, but it sure looks like there is more than one aggregator, for which I have no notion of the number of actual readers.

Could it be that the decelerating growth curve Jon sees in his bloglines sample says more about bloglines than weblog readers? With the increasing competition from desktop readers and other aggregators, perhaps what we are seeing is in fact the deceleration in bloglines growth?

I suppose you could get part of the answer by charting the number of RSS requests from readers used by individuals. Perhaps more new readers are not using aggregators? A bit tricky(!!) to figure out the number of readers from the number of requests, as different folks/products may use different polling intervals.

Also did the recent appearance of Yahoo (and others) support for reading RSS feeds correspond to the drop off in bloglines growth? The timing seems to be about right.

Not that I want to make a career out of this, but it would be interesting to have more accurate numbers.