## Sunday, July 06, 2008

quantcast is an interesting new approach in Internet usage measurement. They have you put a fragment of HTML on the page that you want measured (just like sitemeter or sitecounter do). What's new is that they have instrumented a large number of households in the US. They know quite a bit about the households that have been instrumented. They watch for visits to your web page from these households and report summary statistics to you about this set. (It's a little more complicated than that. Even though they have ~ 1.5 million instrumented households, only a tiny number of these would show up at any blog e.g. this one. Sampling noise would then be unacceptable except for a small number of high traffic websites. They manage to track users across a large number of websites (e.g. this one), and have setup statistical models using which inferences are made. I thank Konrad Feldman of Quantcast for explaining these things to me).

Of course, this is only a measure of the readership of this blog in the US. What it seems to say for this blog:

• 68% are male (a bit more than the average on the net).
• It's a bit above average in the representation of high income groups. 37% have household income from $60k-$100k and 22% are above \$100k.
• In terms of age, 44% are from age 35 to age 49, which is above average. There is of course almost nobody from age 3 to 17.
• In terms of ethnicity, it's 61% caucasian, 8% african american, 24% asian.
• 83% have no child in the household, which is a bit above the overall average.
• 34% of the households have a head of household who's been to grad school: this is 2.5x bigger than the overall average. After all, you do have to be a genius to read this blog.

Watching feedburner, I know there are roughly 2000 subscribers to the feed. Using sitemeter data, there are roughly 16,000 visits a month. Quantcast says these visits come from roughly 7000 unique people who show up at the blog per month. Of these, 2000 are from the US. 52.65% are from India. This number (Indian share) has gradually risen over time: I suppose this reflects the growing usage of the Internet.

You might like to see this piece that I wrote at the end of 2006, looking back at one year of blogging.

Considering that the majority of your blog posts are about India, I am surprised that you dont have a majority of Indian readers.

While I have seen your blog being referenced via other economics blogs such as econlog again that is specific to Indian policy. Are you not surprised by the large caucasian nos.?

Is the age from 35=49 the largest demographic? Who are the rest by age?

2. 61% of the readers in the US are caucasian. That seems to be roughly okay. This describes 2000 of the 7000 people who show up at the blog per month.

Yes, 35-49 is the largest demographic. The really young and the really old don't seem to be interested in this. I guess India as a high growth country that's changing economic policy is a relatively new phenomenon.

More details are on the quantcast site. Are you able to access it?

Please note: Comments are moderated; I will delete comments that misbehave. The rules are as follows. Only civilised conversation is permitted on this blog. Criticising me is perfectly okay; uncivilised language is not. I delete any comment which is spam, has personal attacks against anyone, or uses foul language.