User Behavior Tracking with Google Analytics, Garb, and Vanity
The most successful applications start off with a good idea. From this idea, features and services are created to fulfill the needs of users. Determining how users act when given features has proven to be the best method for guiding feature design. Unfortunately, making this determination is often an expensive challenge, especially if done improperly. This talk will provide you with new tools and techniques to aid gathering information to make these decisions.
With the bulk of the talk I will cover all you will need to know to get information back out of Google Analytics (GA), using Garb to access the API provided by Google. In addition, I will discuss, in-depth techniques and examples for gathering the best data using GA. I will touch briefly on the benefits of A/B testing in order to introduce Vanity. Lastly, I will present techniques for combining data gathered with GA and metrics from Vanity to create a vivid picture of user behavior and how this data might be presented to encourage users. All in all, Google Analytics provides the gateway to a more complete analysis of user behavior, and an invaluable tool for planning the features and growth of your application. Let me show you how to leverage them.
Presentation
Tony Pitale
Viget Labs
Tony Pitale is a developer at Viget Labs, where he builds Ruby on Rails applications for startups and conventional businesses. He’s been developing with Ruby for nearly 4 years and has built web applications with Rails, Sinatra, and DataMapper. He’s the creator of Garb, a Ruby library for accessing Google Analytics data through its API. When not developing he spends time with his dog (Kona), friends in the DC area, and family in NJ.
Comments on this page are now closed.
Comments
Great presentation, But I would have liked to see a little more about what the numbers mean, and what has worked for him in the day to day usings of these tools. And a comparison against google web optimizer.
I’ve added my slides to slideshare: www.slideshare.net/tpitale/...
All of the code from my talk is on github: github.com/tpitale/rc10-cod...
I feel like this talk could have benefitted from a more cohesive “plan of attack”. I’d like to know a solid plan for implementation, along with some options. But the elements were talked about in pieces, making it difficult to walk away with a solid plan.
I also would have liked a walkthrough of code samples. A page of code was sometimes shown, with just a few seconds to read it yourself and try to absorb before the presenter moved on.
I agree with jeff – I feel like I learned what pieces were involved, and now I know what to research.
I really enjoyed your talk. It certainly gave me some ideas for cool features on our apps. a/b testing is still a little obscure for us but with some of the tools that we saw here i think we are more enthusiastic to take on the task.
we already (ab)use google analytics, a/b testing, its time to be able to connect that info with some internal data that we collect from our users. thanks for the insights
Well-structured talk, a little monotone, but great use cases for Garb and Vanity. Now I know what questions to ask when I go learn how to use these tools…
Lots of good stuff here, but it’s presented so fast and in such a fragmented way, that I will just have to go learn everything myself. The code on the screen was too small, but needed more context.
There were some good things w/ regard to finding out about gotchas in GA.