A detailed graphical report of what happens to a site on digg
For the visitors who don’t know, my site was recently featured on digg.com, a website in which users select which stories, images, videos, webpages and such make it to the featured stories list of digg.
This was probably one of the best places to test the digg effect; my site doesn’t usually get many hits at all. I’m used to averaging 4500 unique visitors per month (to those of you who don’t read site stats that’s not really many at all).
Here’s an average day for me on this website (by the way, I’m using Webalizer):
- 500-600 visits (unique IPs*number of times they visited
- 4000-5000 hits (files)
- 1000-2000 page hits
Yes, this is per day. Not too impressive, but then again, I don’t advertise, have much content to get me noticed in search engines, I’m not linked many places, etc etc. I could give you a billion reasons why I’m so unpopular, but I won’t bore you. Truth is, most people get to my site get here by bookmark or a link somebody sends them (direct request accounts for about 30-40% of my visitors in the average month).
The digg effect: a first day’s summary
So, that being said, now we’ll look at how digg has affected my lonely little website. (Note: this is being written about 18 hours since my small budget photography article made it to the front of digg.com):
Sorry I don’t have stats for earlier months. I started on the server I have now in mid-August, so that’s why stats are about half there. Anyway, as you can see, digg has vastly affected my stats.
This is hourly usage over the entire month, but don’t worry about that.. as we determined my site doesn’t really get enough traffic to really affect the digg effect results much. I submitted my digg on 2005-11-07 (and in finding when I submitted I noticed that digg, a techie website, still uses 12 hour time mixed with archaic mm/dd/yy date formats. Amazing.) which was 1 day and 16 hours ago. Around 1100 yesterday (2005-11-08) you can see that my digg made it to the front page of digg.com.
I don’t have any screen shots of this, but I like to monitor my server. Unfortunately I didn’t have any rrdtool graphing software running, so I can’t show you much except my observations. I like to watch my apache logs - I tail them through an ssh terminal and I also watch my server vitals through gkrellm. Here’s what I saw:
|1050||When the traffic really started. At this point my server was doing a constant 100K-300K/s (big K stands for kilobytes). I was noticing my apache logs jumping about three lines at a time (I strip out all the requests but pages, so no, I wasn’t just seeing a bunch of images being requested). CPU is running at about 10-30% by this point (my server has a little duron, btw).|
|1110||Traffic picks up. At this point I was getting worried about whether or not my server would handle the load nicely. It was serving at about a constant 300-500K/s. I was also happy to see that my logs were showing only 200 and 301 HTTPD codes and no 404s. CPU hits 30-50% constant|
|1116||Traffic still going up. CPU hits 100% for about two seconds and freaks me out|
|1147||Traffic picks up even more; we’re doing about 500-1000K/s constantly. CPU is at 30-70%.|
|1450||Traffic maintains its flow, my digg makes it to the bottom of digg’s front page|
|1457||I leave digg’s front page with 410 diggs|
For the next few hours I began noticing a lot more stuff. I was getting hits from digg.com/topstories, del.icio.us and of course digg page2, page3 and so on. At this point traffic didn’t really decline all that much, but it did noticably enough.
Here’s the small list of referrers:
The blanked out line is referrer spam. I just learned a few methods to block this a few days ago.
This was definitely quite a different day for my server.
My article had a ton of links in it. In the first paragraph there was a link to my gallery and my before and after page. These were both getting a lot of hits and then of course hits off of there to my individual gallery pieces. The small budget photo page was also generating a lot of new page traffic from all the thumbnail examples that people were clicking on. Occasionally someone would click on some of the menu items, but it was rare.
- Almost all diggers are FireFox users (much to my dissapointment.. I’m an Opera user. Definitely happy I didn’t see a bunch of IE though). I saw mostly 1.0.7, then 1.5. Opera people were all running 8.5
- Quite honestly there wasn’t that much traffic coming from digg. Had my page just been a page without all the links and thumbnails and people actually going through the rest of my site I probably wouldn’tve seen much load at all on my server
- I’m proud of my little 512mb, 1300 Duron server
- Lots of comments, but not a big deal to moderate them all. You definitely want to be able to moderate them though.
- Didn’t notice much real blatant flaming or spamming
- Diggers do not click on ads
- A lot of traffic comes from other sites besides digg. I’ve been getting a ton of hits from other web logs and community linking sites like digg.
Site related and personal stuff
You really need some balls if you’re going to put anything up for people to see. Comments aren’t always the nicest. A lot of people were offended by my lens comments in the cameras section of my article, but I figure that came from A. The people that didn’t understand that I was trying to say one good lens and even one good fixed lens on a non-SLR would work fine or B. The people that were the ones who bought a ton of lenses so they could feel and appear professional and didn’t want me ruining their fantasy.
I found a lot of people who did and didn’t like my style. Lots of comments about my high-saturation/high-contrast work. Well folks, it’s my style.. and if you don’t like it you can get the fuck out.
Got a lot of comments about Sasha in the leaves picture. Don’t have much to say to that. I try to protect the identities of my models and so far I’ve had no problems =) But I’ll let her know how much she’s appreciated ;)
There was a lot of positive feedback too, moreso than the negative. You may not be able to see it in the comments so well, but just watching my apache logs I can tell you that a *ton* of diggers or diggers who sent their friends to my site looked through almost the entire gallery. Even a lot of the negative feedback was really positive; there was a lot of creative criticism and it will affect how I do things in the future.
If your site is preparing to be dugg I suggest the following:
- Don’t expect revenue
- Expect high traffic and make sure your sever or your account can handle it. I generated about 1gb of traffic since when I got on the front page till now (it’s only been about two days). I usually get about 50-200mb of traffic a day.
- Make sure if you have a comments system or forums or anything that you can easily moderate it
- Try to find a way you can easily watch your server. It really is a lot of fun to just sit here and watch the apache logs trail by and look at server health through gkrellm, especially if your website isn’t usually all that busy. I really suggest you get some monitoring tools for your server so that you can look at things more easily during and after you get dugg - I really wish I had something installed while this happened.
- Check all your links and make sure they’re not broken. This is the one time when you really don’t want anyone to get 404′d.
- Grow some balls and prepare for feedback. It was great hearing all the different thoughts on my article and work. I even got some emails and instant messages with comments, questions and even a job offer. If you want this make sure you have a comments system that’s easy to use or have an easy way to contact you.
It was really nice to be dugg. I liked being able to share some information with people who wouldn’t normally see it. I didn’t mind not getting any ad hits off of the users since I really was just hoping for a bit of exposure and the pleasure of someone actually appreciating that article I wrote since it took a bit of work.
I’ll probably make a followup to this later, but quite honestly traffic has really died off; it’s still much denser than normal, but almost not worth mentioning after all that excitement of yesterday. I expect a few more gigs worth of traffic because of this, but nothing really sustained over a long period of time. If that changes I’ll definitely let you know though. Observation right now is that I’m not really getting many hits from digg at all and I get the feeling most of the traffic is people sending each other links, so I’m getting even more direct request traffic than normal.
Overall this was a really rewarding experience and I’m honored to be dugg by so many people, including Kevin Rose (also kevinrose.com). I’d also like to thank Josh Kearney - if it weren’t for him I’d probably never be a digger, Matt Sparks for helping me get all my server stuff up and running properly so that I could share this wonderful information with you and all of you who made this possible by coming here, visiting and letting me know what you thought =)