I was lucky enough to have my Decline of Facebook article make the front page at Hacker News on April 4, 2011, the day the article was published. I wanted to drop some numbers here and show the results of having that much traffic hit my site. These things interest me, so I’m sure many of you will be interested as well.
Here are some details to put the numbers into context. The article reached a max position (as far as I know) of about #14 on the front page and averaged around position #18. It was listed on the front page for about 1.5 hours. Not a huge amount of time or a super high ranking, but it still provided a tremendous amount of traffic.
Basic Analytics
The article received about 2755 visits on April 4 and around 343 on April 5. So let’s say just 3100 total pageviews in 2 days. According to my analytics about 95% of the pageviews can be attributed to Hacker News, either directly or indirectly.
- Of those 3100 pageviews, only 186 readers actually looked at other articles on the blog (1.06 pageviews per visitor, 95.87% bounce rate).
- Average time on the site was 4 minutes 10 seconds, plenty of time to read the entire article.
- Nearly half of the traffic coming from HN shows up in Google Analytics as “direct” traffic. For whatever reason GA isn’t able to properly identify that traffic as being from HN, but I know that’s the source.
Server Impact
I use and love ChartBeat, so when I get significant traffic on any site I stay glued to my monitor to “watch” as users move around the site. During the 1.5 hours I was on the front page at HN I consistently had around 200-220 visitors on the site at any given time. I use Rackspace Cloud for my hosting, so I noticed no impact on page load times.
My blog loads a little slowly due to the graphics, lengthy text, etc. It varies between 2 seconds and 3.5 seconds, depending on the page. I realize that I need to implement some more caching on my WordPress install, but I know that server strain wasn’t an issue because user page load times were the same whether there were 2 users on the page or 200. Being able to monitor user load times is a key feature that I use from ChartBeat.
Lasting Impact
I have 21 new subscribers to my RSS feed that I can attribute to the HN article. That equates to about 0.7% subscription rate. It’s harder to tell about how many new Twitter followers I received as a result, but it’s negligible, if any. My last count had about 30 tweets with links to the post.
This was far and away the most popular article that I’ve had in my 2 months of blogging. Thanks to all who read and took the time to make comments or share the story.