Webmaster's Blog

The academic year and the state budget year both run off kilter from the calendar year. So what a normal person would consider the start of a new year is mid year for us. One thing a lot of us do is compare where we are to where we hope to be by year’s end. How formal the goals and plans and data involved in all that is can vary a lot. I don’t have much in the way of solid goals. I simply strive to improve. But data I’ve got.

Let’s take a look at some general performance metrics for our website. I’ll even try to supply a visual aid. I have no idea how this is going to look when I’m done. I’m used to hand coding all this stuff.

Ok, if you’re seeing what I just saw when I checked the preview, this will work. Both the tiny images above link to full sized versions. Feel free to refer to them as needed.

What are we looking at and what is being compared?

Those are timelines from Google Analytics comparing total visitors for the first 11 weeks of the redesigned site (Monday November 2nd 2009 thru Sunday January 17th 2010) to some other 11 week point in the past. In both cases, the thicker blue line is the new site and the thinner green line is the old site.

The graph on the left compares to the same time frame a year ago (Monday November 9th 2008 thru Sunday January 18th 2009). The graph on the right looks at the 11 weeks leading up to the launch of the new site (Monday August 17th 2009 thru Sunday November 1st 2009). There are pros and cons of each view.

Apples to Oranges?

Comparing figures from a year ago will inflate some things because our traffic is higher for all the same reasons we’ve got a significant enrollment increase this year. However, if you look at that graph full sized you’ll see the general trend of peaks and valleys matches up because both data sets include the same holidays and what not. The 11 weeks leading up to the redesign launch corrects for the general enrollment increase inflation effect, but we’re now comparing the busiest time of the year (leading into Fall semester) to the slowest (winter break). If you look at that graph full sized, you’ll see only the valley’s on the weekends match up and the other trends are all over the place.

I wouldn’t go so far as to call either case comparing apples to oranges, but neither case is perfect either. We’ve got Galla apples and we’re comparing them to Fujis and Granny Smiths.

The Numbers

Definition of terms

1. Visits: Someone visits a page on our site and moves on to at least one other page on our site before surfing to someone else’s site. That’s one visit. They may come back later for another visit. We generally want this number to go up.
2. Unique Visitors: A single unique visitor may produce dozens of visits (#1). This is another one where moving up is good.
3. Pageviews: The unique visitors (#2) drop by our site for one more more visits (#1) each consisting of 2 or more pageviews. This number also reflects sessions where people just see one of our pages and immediately leave. Those aren’t visits, those are bounces (#6). This is a number we like to see go up, but we need to put it in context of the first 2 items on this list to make sense of it.
4. Average Pageviews: Total pageviews (#3) divided by total sessions. Think of sessions as the total of all visits (#1) plus all bounces (#6). This is a good indicator of engagement. The more pages a person looks at on our site, the more interested that person is in the information we are providing them.
5. Average Time on Site: Another strong indicator of engagement. We generally assume someone who spends 15 minutes on our site is more interested in what they are finding there than someone who spends 2 minutes on our site. I believe this ignores bounces (#6) and is calculated per visit (#1). It also ignores any single pageview (#3) that lasts longer than 20 minutes assuming that someone left a page open but wasn’t actively reading it all that time.
6. Bounce Rate: This is the only number on this list we want to see go down. This is when someone sees one of our pages and leaves without looking at any other pages on our site. This is generally considered a bad thing, although it could indicate that person found exactly what they were looking for on that first page such as a link to get into their email or to D2L. This is reported as a percentage of total sessions.
7. New Visits: The percentage of total visits (#1) from people who had never been to our site before (at least not as far as Google can tell).

I’ll put everything in terms of percentage change so that it makes sense without having to understand all the details. The numbers below should be fairly clear even if the definitions above leave you feeling confused.

Compared to 1 year ago

• Visits up 31.61%
• Unique Visitors up 78.14%
• Pageviews up 33.74%
• Average Pageviews up 1.62%
• Average Time on Site up 14.97%
• Bounce Rate down 37.46%
• New Visits up 56.17%

Compared to time leading up to redesign launch

• Visits down 11.49%
• Unique Visitors up 6.0%
• Pageviews up 6.65%
• Average Pageviews up 20.49%
• Average Time on Site up 25.19%
• Bounce Rate down 45.98%
• New Visits up 30.17%

Obviously we’re dealing with a lot of variables unrelated to the redesign of the site. But between these 2 sets of data you can get a pretty good idea of our overall site performance. I’m most excited about the metrics related to engagement. Even now that most of the “new” has had time to wear off people seem to be spending more time getting to know us through our website.