Hacker News Reader

Comment by pg | Hacker News Reader

This reminds me of the article posted here recently about the locations of ancient Woolworth's stores. If you'd written an article 3 months ago that led off with the compete.com graph of the two sites' traffic, you could have written about how Posterous was kicking Tumblr's ass, and used that to support an equally plausible sounding thesis about how this was because they were based in Silicon Valley.

This article is like an inverted pyramid, balanced on a couple months' trend in compete.com stats. Not actual traffic stats even, but compete.com estimates of traffic, which as anyone who runs a site knows, can be way off.

View on HN

From Compete - yearly change:

tumblr: +184.36%

posterous: +297.36%

So, uh, even with the last 3 months in posterous is still growing faster than tumblr. If it were a stock its valuation would have increased more than tumblr. Also, wouldn't there definitely be less hits to posterous based on the fact that you EMAIL posts in rather than sign in? That being said, the tumblr landing experience is probably one of the reasons I blog with them and not posterous.

Great points. I'll incorporate them into the post.

It's growing faster from a smaller base, though. I find more relevant that Tumblr is still much, much bigger. And I think it will stay that way.

One way to normalize this is to ask: How quickly was Tumblr growing when it was as small as Posterous is now? Then you're comparing apples to apples.

Thanks for the comment.

If you have actual traffic stats I'd love to see them. ;)

Also, the estimate isn't for "a couple months", but for the past 12 months, and the trend points in one direction. Even less reliable but even more "damning" is Google Trends: http://google.com/trends?q=tumblr,+posterous&ctab=0&...

That being said, your general point is right: nothing is preordained, and the trends could switch around tomorrow, and the picture look much different a while from now. But that's why I have a blog: this is the situation as I see it today. Maybe six months from now I'll eat my words, who knows?

The couple months I was referring to are the last few. The preceding months contradict your thesis.

I think you're mistaking how much license a blog gives you to write about "the situation as I see it today." It's ok to write stuff based on small amounts of data, but the conclusions you're entitled to draw from it should be correspondingly small.

They don't contradict my thesis since Tumblr was always much bigger than Posterous and while they may have had faster growth it was never enough to catch up.

Again -- I may be proven wrong in six months or a year, and if so I admit it publicly right here (in fact, do you want to make a bet?). But I think at this stage the data warrants the trend I extrapolate from it.

compete.com is a joke.

look at the top search terms sending traffic to each site. they include things like "swine flu" and "cleavage lover", which IMHO are indicative of a statistically insignificant sample size

Reply on news.ycombinator.com