As most of you probably had guessed, I not only enjoy typing little gems of semi-worthless knowledge on this site, but I also enjoy doing a little data mining on some of the statistics every now and again.
So, this evening, we’re going to take a little trip into the statistics of HandlebarSandwich.com and see what we can dig up. I monitor the statistics of the site pretty regularly, but I figured that some of you may be wondering to yourselves “am I the only one who reads this stuff?”. Stay tuned, I’ll give you more than you ever wanted to know about your fellow visitors. . .
Tools of the Trade
The first item of business is the “how” for the stats. I primarily use 3 different tools to track stats on the site. The first is Google Analytics. This is where I get all the general visitor information including where people come from, how long they spend on the site, what they look at, etc. etc. The second is Feedburner. This is very specific to those of you visiting through a RSS reader, and tracks all the feeds coming in and out of the site. The final one that I like to check is Technorati. Technorati allows me to see who is linking here from other blogs. It gives a nice little list of who is quoting me and tells other blog users that I quoted them.
Note: Unless I state otherwise, I’m going to be running down the stats from January 1, 2008 to May 4, 2008.
Who?
The first and foremost question always comes down to “How many people visit your site?”.
According to Google, since 1-1-08, I’ve had 5,898 visits, which comes out to an average of 46.81 visits a day. This is a little bit skewed because it doesn’t count the people who check my site through an rss reader and don’t click through to the post. On the average, I have 42 people following the site through a RSS reader.
As we dig a bit further into the google data, I took a look at the overall graph of visitors, which looks something like this:
As you can see there is definitely a repeating pattern to the data, so I got a bit curious and decided to download the raw data and do a bit of digging. I plotted the number of visitors by day of the week and got a chart like this:
It certainly looks like a pattern. Turns out that the averages come out looking like this:
Monday – 60.3
Tuesday – 55.8
Wednesday – 52.1
Thursday – 50.5
Friday – 40.9
Sunday – 39.2
Saturday – 26.4
I’m happy to see that everyone gets the monday blues and reads my site. I’m also happy to see that people are out doing something other than staring at a computer over the weekend.
Geek Extra Credit: Not quite satisfied with just looking at the numbers, I dumped all the data into a statistics program that I use for work (JMP) and ran an ANOVA to see if there was a statistically significant difference. For those of you who are actually still with me on the extra credit, the p-value of the ANOVA was <0.0001. A Tukey-Kramer analysis yielded the following chart, showing which days were statistically similar (if they share the same letter, they are similar), and which ones were statistically different.
After digging into the weekday phenomenon, it only follows that time of day should also make a difference. So, here it is:
Where?
Where means two things to me. First, where are you physically, and second, where did you come from to get to my site.
As far as physical location, you can imagine that the majority of you guys are from the US. Here’s the breakdown of the top 5 countries with number of visits since january 1:
US - 5654
UK - 48
Canada – 46
Germany – 23
Japan – 14
After that, it’s also no surprise that the majority of you read from Arizona, although some other places are picking up steam:
Arizona – 2209
Colorado – 704
New Mexico – 366
California – 281
Indiana – 227
Next is another important question, where did you come from on the internet. Again, no surprise that the majority are direct requests, but the new frontrunner of late is Mr. Gnome himself from onespeeder.com. If you had checked this list about a month ago, Big Johnny was in the lead without contest, and now a lot of you are linking over from Onespeeder.com and tRoy’s site at epicrider.com. Very interesting. . .
(Direct Request) – 2305
Google – 626
Onespeeder.com – 489
Drunkcyclist.com – 480
juniper-solutions.com (which covers tRoy and dara) – 396
Forty15.com – 160
Keywords
I am always amused as to what people search for to find my site, and this definitely gets one of the ‘most surprising’ awards.
“handlebar sandwich” – 168
“handbar sandwich” – 50
“fixed gear handlebars” – 14
“fkr, kevin rice” – 12
“surly big dummy” – 11
Yes, you read that correctly. I had 12 people find my site from a search for “fkr, kevin rice”. In fact, if you go to the google homepage, type in “fkr, kevin rice” (no quotes) and hit “I’m feeling Lucky” you will end up right back here. Bitchin.
Pages
So what are people looking at the most? Obviously, the main page gets the most views by far, but since january 1, here are some more top pages:
Gallery – Sedona Circ 2008
Steep and Cheap, a love-hate Relationship
Soapbox – Customer Service 101
Thinking about a Big Dummy
Gallery – Winter Pursuits
RIP Sheldon Brown
That was from google analytics, if we go in from feedburner, we get an idea of what people are clicking through to from the RSS feeds:
Getting off my ass – part 2
Road Ride and a Training Confession
Road Trip
Win Free Stuff or Buy Cool Shirts
Regaining Motivation
It’s really interesting that people click through to different posts from a feedreader versus on the site. Not sure exactly what to make of that.
Final Thoughts
That’s about all I have from the stats for the site. There are more out there to look through, but I think those are kinda the most interesting and telling ones.
The only other one I will point out is my technorati “authority” score. This is a measure of how many people link to your site from other blogs. Right now I’m hovering around 12. To give you a basis for comparison, onespeeder.com is at 8, drunkcyclist.com is at 98, fatcyclist.com comes in at 293, and bike snob NYC comes in at a whopping 542.
Finally I have one more bit of geek extra credit. We got a demo on the new version of JMP at work the other day, and it has this feature called “animated bubble plots”. Basically, it allows you to visualize how data changes over time. The instructor joked that he didn’t know what you would use it for, and so I aim to prove him wrong.
There are several things going on on this plot, and here’s what they all mean:
- the x-axis is day of the week
- the y-axis is number of visits per day
- the size of the bubbles represents the average amount of time spent on the site on that day
- the color of the bubbles represents how many pages, on average, a person viewed when they visited the site that day (red is more than blue)
- the time factor is displaying weeks of the year since January 1
It’s easiest to run through the video and focus on one or two things at a time, which can help to see the different relationships between the factors.
My head hurts.