I have to say that I really enjoyed Kaiser Fung’s presentation last night. I probably asked too many questions as it was but I have always been amazed with data and the story it can tell. In my position as a Web Metrics and Data Analyst, I am responsible for the analysis of all the data that comes through ETS.org. Within this domain we have over 50 different products, all of which the data means different things. For instance, in our Research section, success is measured by how many papers are downloaded and how many unique visitors come to the site. In regards to our tests, registrations is the single most important goal we want visitors to complete when visiting the site. With Social Media now become an integral part of our marketing plan, measuring the effect it has on our web traffic has become increasingly more important. To do that, it is imperative to know where your traffic is coming from.
First off it is important to note that within analytics programs such as Google Analytics and WebTrends, you can always see where your traffic is coming from.
This will allow the analyst to see what websites are linking to your site and in what context.
With the rise in usage of short URLs, many people are using them in their posts. This allows the poster to see how many people click on their links (and with enterprise editions, more data). Unfortunately the following scenario for the referring page to your site now becomes this:
Posters can now link to your site, without leaving a trail to follow. In social media that can be devastating. Below is an example of a post on Twitter:
As you can see, even though there is a URL in there, Twitter still inserts their own URL. When you view the referral data in Google Analytics, this is what you see:
If you are personally placing these posts, you can use parameter tracking to track the success of your posts/advertisements but that only accounts for some of your traffic. There are advanced techniques that you can utilize to identify where these short URLs are being placed by 3rd parties such as:
- Backlink Checkers
- Spidering third party sites
- Using search engines to identify where the links were placed
- A lucky search on Google with implied knowledge of where the short URL was linking to
These methods are cumbersome and labor-intensive, usually with little to no reward when weighted against the amount of time spent on each link. When you do identify them, you then have to line up the data with the appropriate URL in your analytics program and then measure that against which product it is about. As mentioned earlier, this can be even more labor-intensive given that different products has different goals.
I think as a Web Metrics and Data Analyst, you have to pick and choose what short URLs you want to investigate. For the most part the single biggest factor would be the amount of traffic each URL sends your site.
As a user, I love being anonymous. As a Web Analyst, I loathe it.