![this is why i love matt langer.
langer:
So I had a little fun with my latest project at work.
Every business realizes at some point that Google Analytics sucks but gets away with sucking because it’s Google and it’s free. We were looking to get a better view of our viral-specific traffic, and rather than wrestle with GA or try to plug in to another service we decided to farm our own data. It was up to me to figure out how best to do this.
My first step was to build our own vanity URL shortener, functionality we would definitely need because Twitter is one of our three main viral channels. I wasn’t comfortable with plugging into the API at bit.ly or anywhere else because I didn’t want to have to rely on another service for the longevity of our links. Plus, by doing it myself I not only got to build my own base-62 converter (fun!), but now whenever a user hits our server to expand that base-62 shortened URL I get full server-side control of the rest of that click’s lifetime. I know if that URL went out via Twitter or Facebook or email, I know its incoming referrer (if it differs from the original channel), and most importantly, I know the unique user session of whoever clicked on it.
Now, with a unique session in hand, I built a JavaScript tracker to capture a complete view of that first user experience. I record the entire map of their initial visit, including which site features they used, whether they created an account, and about a dozen other properties that interest us, all of which builds out a very important picture of customer acquisition and conversion. It also helps us gauge the respective value of these three separate viral media and track their performance over time as we tweak the ways we interact with them.
I was also asked to build a live dashboard so that we could hang a 50-inch LCD on one of the walls and get a realtime view of our incoming viral traffic. And while I find analytics and metrics and viral coefficients to be pretty interesting stuff, my cynical half can’t help but see it as all a lot of snake oil. In deference to this latter point I decided to name the tool the “VIRAL SCIENCE LABORATORY”, since I hear “viral scientists” are the new “social media experts”.
They don’t usually let me do much design around here. This might be why.
(full version here)
—
[ed note: the numbers pictured above are just my local testing data, not a representative view]](http://27.media.tumblr.com/yvZqiroI9q992uwcj51hvHt9o1_500.jpg)
this is why i love matt langer.
So I had a little fun with my latest project at work.
Every business realizes at some point that Google Analytics sucks but gets away with sucking because it’s Google and it’s free. We were looking to get a better view of our viral-specific traffic, and rather than wrestle with GA or try to plug in to another service we decided to farm our own data. It was up to me to figure out how best to do this.
My first step was to build our own vanity URL shortener, functionality we would definitely need because Twitter is one of our three main viral channels. I wasn’t comfortable with plugging into the API at bit.ly or anywhere else because I didn’t want to have to rely on another service for the longevity of our links. Plus, by doing it myself I not only got to build my own base-62 converter (fun!), but now whenever a user hits our server to expand that base-62 shortened URL I get full server-side control of the rest of that click’s lifetime. I know if that URL went out via Twitter or Facebook or email, I know its incoming referrer (if it differs from the original channel), and most importantly, I know the unique user session of whoever clicked on it.
Now, with a unique session in hand, I built a JavaScript tracker to capture a complete view of that first user experience. I record the entire map of their initial visit, including which site features they used, whether they created an account, and about a dozen other properties that interest us, all of which builds out a very important picture of customer acquisition and conversion. It also helps us gauge the respective value of these three separate viral media and track their performance over time as we tweak the ways we interact with them.
I was also asked to build a live dashboard so that we could hang a 50-inch LCD on one of the walls and get a realtime view of our incoming viral traffic. And while I find analytics and metrics and viral coefficients to be pretty interesting stuff, my cynical half can’t help but see it as all a lot of snake oil. In deference to this latter point I decided to name the tool the “VIRAL SCIENCE LABORATORY”, since I hear “viral scientists” are the new “social media experts”.
They don’t usually let me do much design around here. This might be why.
(full version here)
—
[ed note: the numbers pictured above are just my local testing data, not a representative view]
