Understanding Web Statistics
Posted: October 14th, 2008 | Author: Bryan | Filed under: Analytics |If you're new to Mid Mo Design, you may want to subscribe to our RSS feed or contact us regarding any assistance you may need. Thanks for visiting!
Web statistics packages are now mainstream and very much accessible to the average web administrator; in the past, this wasn’t always the case. When Google bought Urchin in April of 2005, it was already an industry leading company specializing in providing an statistical package to webmasters allowing them to track pageviews, clickthroughs, sales, bounce rates and more for each individual user. Now, it is the number one analytics package available online, it is free, and it is extremely popular.
So let’s learn a little about web analytics. While it is a great tool and gives insight into your most important of commodities, your users, it has its limitations.
Basic Statistics
There are several statistics available with even the most rudimentary analytics program. These include page views, visits, pages per visit, bounce rate, average time on site, and percentage of new visitors. Let’s break these down:
Page Views
This stat simply tells you how many pages were served in the time period, it is the most basic statistic and does not impart a lot of information about your website. However, it is great for understanding the scale of your hardware needs, especially in conjunction with load monitoring on your server itself.
Also, it is a great indication of the number of ad impressions you are serving. If you are an advertising website, you want this number high to increase revenue.
Visits
This stat lets you know how many times your website has been called upon at least once in a given time. This statistic is a bit more advanced than page views because it keeps track of sessions. For example, if a user looks at 5 pages now, leaves and returns 20 minutes later to double check a single page, your analytics package will rack up 2 visits and 6 page views.
Pages Per Visit
Pages Per Visitor tells you how many pages (on average) a single visitor viewed. A high number means that your visitors tend to view several pages, which is good.
Bounce Rate
Bounce rate is a bit more advanced and definitely more useful. It tells you what percentage of visitors arrive at your website and leave before viewing extra pages. You want this number LOW, unless for some reason you have a website with only one page. If you only have one page, its impossible to get that second click. This is an extremely important stat, more on that later.
Average Time on Site
This lets you know how long your average visitor stays on your website, this is usually highly correlated with Pages Per Visit; the more pages they visit, the longer they generally stick around.
Percentage Of New Visitors
This is a great stat for letting you know the degree of influx of new visitors. Alone, it just gives you a general outline of the mix of returning and new visitors but without some more information, like growth in visits, it is just a number.
Basic Trending and Forecasting
A great way to use your new found statistics is for trending. You can get an estimation of growth or how your user mix changes over time. For example, a young website will have a new visitor stat in the 90% range while a mature blog will likely dip under 30% or 40%. So a little context is required to make sense of your site’s growth.
A major setback in Google’s Analytics package is the fact that it lacks a forecasting feature. To top things off, it also doesn’t have a robust export function for friendly CSV formats. In a perfect world, your data would be free to analyze and forecast on your own, but I digress.
The most common way to trend growth is with a mix of Page Views and Visits. Using Excel, one can generally graph and project growth with either a linear or exponential curve. Linear curves are more realistic as exponential graphs tend to be great in the short run but highly optomistic in the long run.
Website Optimization
Another excellent use of analytics packages is the ability to optimize your site. Here is a quick solution for a site with an abnormally high bounce rate:
The best way to deal with a high bounce rate is to drill down into your content. Check out which content pages are the worst offenders. Try to identify why the pages with a high bounce rate differ from pages with low bounce rates. Try adding something to entice the user into making a second click and compare before and after. Did it work? Is the bounce rate lower? You can do this indefinitely.
Another way that analytics packages have transformed the face of the web is with the use of goal funnels. Goal funnel are customizable and track each page leading to a sale. For example, clicking a products page might be the first step. Clicking a buy now might be the second step. Entering your username and password or creating an account might be the third step. Entering your contact information might be the fourth step and so on until the item is purchased and the user is directed to the final step, a thank you page!
By keeping track of how many steps a user takes before changing their mind, you can identify problem areas in your funnel. Asking for too much information to sign up? Do you really need two phone numbers? Try correcting that and see if your conversions increase.
Get Involved
The best way to get a feel for analytics is to ask your tech guy for assess to the web stats. It is likely that he or she will be more than happy to let you view the statistics for your website. At Mid Mo Design, we employ industry leading analytics packages in conjunction with hands on analysis and optimization. Every project we tackle comes with a unique analytics package and the client is encouraged to become engaged in tracking our success. Contact us today if you need assistance with any of the above.
Leave a Reply