I never thought I could be as obsessed about data as I have become over the past months even after spending 6 years at a data obsessed company. EC and I constantly pour over our reporting and have been aggressively implementing tracking to understand the dynamics of our site, users and acquisition channels. I have to admit EC is a little more obsessed than I am – he looks at the logs and Google Analytics constantly and loves to give me up to the hour reports (and even up to the minute if we are close to hitting a milestone).
So there are two reasons why I thought I would blog about reporting…
1. Building a business: After talking with some entrepreneur friends I’ve come to the opinion (accurate or not) that if you aren’t obsessed about the metrics of your start-up then maybe you need to rethink whether you have the right personality to build a business from the ground-up (this of course assumes that you are one of the main guys/girls of your company). I’m not making any judgement calls here; I’ve just found that getting your start-up off the ground and working to turn it into a real business (with lots of users) requires an obsession with metrics. You need to know what’s working or not (with the product, monetization, marketing, traffic sources, etc.) and iterate intelligently and quicklt on all of these “programs”. This requires a different mindset than when you are in the actual process of developing your product.
2. Google Analytics: Google is a great free product and a great product for starting your company but it has its limits. For transparency, I became intimately involved in the Omniture product (the enterprise version of Google Analytics) while at eBay – I kind of miss it but it’s super $$$. Here are some things we learned after taking quite a bit of time to implement and use Google Analytics. Thought this might be helpful for others (and hopefully I am correct!).
a. That funky 0:00 Average Time on Page.
- I’m sure you’ve seen this and said “huh!?” In order to calculate time on a page, Google needs both a T1 and T2. T1 is the time when you enter a page and T2 is when you go to another page on that site. When you go to the next page, that page pings Google and provides it with a T2. The issue is if a visitor only visits one page on your site (recorded as a Bounce), Google can never get a T2 for your visitor.
- Implications: If you look at the Content reports, the Average Time on a Page summary EXCLUDES 0:00; however, if you look at Average Time on Site, this value INCLUDES 0:00. So this will likely mean that your Average Time on Site will be less than your Average Time on Page. The differences in values can definitely make you scratch your head. My next step is to see if I can easily get the Average Time on Site when Bounces are excluded – I am not confident that this will be easy to do.
b. Event tracking and bounce rates
If you implement event tracking be aware that your bounce rates will decrease. On pages where event tracking is implemented, the page pings Google effectively giving it a T2 mentioned above. If you are going to implement a lot of event tracking I recommend that you take the time to map out how you are naming your Category, Action and Label and to understand how your site has been created. We have tracking at the .jsp and tag level so this created a little bit of complexity. ps – we had some issues implementing event tracking (EC said the Google instructions were wrong). I don’t have the exact details but if you are having some issues even though you believe you are following the instructions let us know and we’ll see how we can help.
c. Path analysis
Oh I do so miss Omniture for this! Unfortunately you cannot see, in aggregate, how people got to a group of pages and what they did after they left that family of pages. You can only look at how people entered and exited a specific page. For BlazingStreams we have “Article Pages” – I would love to see how, in general, people got to our article pages and what they did next but unfortunately I can only see this by clicking on a unique article. This type of reporting becomes valuable when, for instance, you can identify page types that drive additional click-thrus or page types that are particularly popular with certain traffic sources. Our site is very dynamic so looking at a specific page (unless it’s the Homepage/static) isn’t very useful.
d. Advanced Segments
After trying this, I’m not sure it’s going to be that useful for me beyond the basics. The important thing I learned is that Advanced Segments does not function like the filters in the Content section. My goal was to understand which categories of content people are most interested in (Sports, Health, etc.) and from this create segments based on Category type and track their behavior as a group.
For example, based on our website model this means I want any page associated with a Category ID I am interested. For example, for Sports I want any Category, Topic or Article page associated with our Sports ID. When I entered a request to pull results that contained our Sports IDs, Google Advanced Segments pulled every page related to a visit that viewed at least one page with the Sports ID. So if EC looked at an Article Page associated to our Sports ID, Google Advanced Segments will also show me that he looked at the Homepage and any other page during his visit. So in order to get the data I’m seeking, I am now filtering the information from Content and recording it in Excel (not ideal but you gotta do it).
I’ll post more info about reporting that might be useful for others as we come across it. If I’m thinking about my reporting wrong – let me know!
Cheers!
), we’ve been working to make BlazingStreams easier and more fun to use, and the content more relevant to you.
sign located next to each topic or category throughout our site to add it to your BlazingStreams homepage.
sign.
