On: Reddit's traffic grew by 300% in 2010

[vladd]

Traffic in December exhibits weird patterns (lower for business sites and higher for leisure sites due to the holidays).

I wonder if they avoided a December 2009 to December 2010 comparison on purpose (or if they preferred the Jan-vs-Dec one to get a better growth).

[jedberg → vladd]

They would have been more impressive if we did Dec to Dec. Our traffic is usually down in Dec.

We did it that way because we were reviewing 2010.

[swores → vladd]

Not quite as simple as business vs. leisure, but I agree that the only reason they'd chose Dec10/Jan10 not Dec10/Dec09 is to make the stats look more impressive.

[jedberg → swores]

Actually, they would have been more impressive if we did Dec to Dec. Our traffic is usually down in Dec.

We did it that way because we were reviewing 2010.

[swores → jedberg]

Then that's really odd, the way any business would normally review a year's performance is to compare December/December, because that's the best way to get a decent comparison without being influenced by factors such as "Our traffic is usually down in Dec." Same thing for other time periods - if you want to see how good traffic was over a weekend, you don't compare Sat/Sun to Thur/Fri, you compare Sat/Sun to the previous week's Sat/Sun.

  Our traffic is usually down in Dec.

Oddly, I've found that to be the case every year, across a range of non-business for-leisure websites.

[jedberg → swores]

We weren't trying to review our performance year to year though. We were just showing how the beginning of 2010 compared to the end.

All the media decided to make it a year over year comparison, which it was never intended as.