Continuing our analysis, we want to get a better measure of the annual growth. The first column takes the blog popularity from the 12 months ending 9/17/2005, and compares it to the prior 12-month period. As noted before, we are using the odd cutoff date of 9/17 to roughly correspond to the TimesSelect period (see data). 

Each column is color-coded to give a sense as to the rate of growth. The initial spurt of "hypergrowth" is statistically meaningless since the first year of data was so small. This is mostly due, as mentioned, to Google not indexing past entries in a blog once discovered.

The general trend is that the rate of growth has been slowing each year.

The NYT columnists, and 4 TimesSelect bloggers, are noted in bold below.

(Note that 1.3 is the same as 30% growth, but to write it that way would necessitate writing 92.7 as 9170% growth.)

The average grew by 33 in the first year; 3.7 times in the second year; and 1.8 times in the last year. 

Some of the bumps in the data can be explained by our sense of real world media events. Robert Novak's popularity tied for the slowest growth in the second year– a "mere" 80%– due to his huge notoriety the previous year from his role in the Valerie Plame leak investigation (as for Anne Applebaum's drop, no explanation comes to mind right now.) Joe Klein, Fred Barnes, and Jonathan Alter all had a nice 10x boost in 2006; perhaps due to the publicity surrounding their books. The irony is that, by authoring books, they weren't obviously giving away the content; clearly one can whip up blog buzz through traditional marketing (though there were likely excerpts and interviews which provided fodder for blogs).

Four of the Times' big 7 rebounded in the second TimesSelect year relative to the average; Maureen Dowd, David Brooks, and Nicholas Kristof fell short. It may be due to changing tastes; I don't know. (It will partially be addressed in Part 6). 

The next page will illustrate the data as absolute numbers using a graph.