Badgeville’s 5 Top Analytics Principles

Following up to my earlier post introducing the basic concepts of analytics (Analytics 101), I want to share with you how the Badgeville team views quantifying and measuring the health of your online experiences.

Badgeville’s fundamental analytics principles are:

1. Keep the user interface for viewing and understanding your analytics as simple as possible! Only display the metrics that matter, and keep them logically organized.
2. Store as much raw data (such as metadata around a particular behavior) as we can at the lowest level possible (per player, per activity) to keep our options open for future insights.
3. Don’t separate analytics into internal, for Badgeville only, and client-facing. Keep everyone on the same page. The analytics we show to customers are the same exact ones we are using to measure the our programs.
4. Never build “black box” canned solutions. There’s no magic here — only deliberate and transparent work designed for our customer needs.
5. Stay flexible, learn and adjust as we go. There’s no one-size-fits-all analytics solution, and our job is never done.

Analytics is a complex field with numerous possibilities to compute and provide an intelligent layer to data. What does that mean in terms of building the right data processing and visualization engine? It means identifying what’s most important and focusing on it. How? By talking to everyone involved, both internally and externally.

Analytics is empty without understanding the consumer. Good analytics is not about recording and measuring everything and producing numerous and complicated reports and dashboards. It’s about narrowing in on what really drives the business and surfacing it in the best, most easily digestible way possible. And what’s important varies by business unit and each individual client. One may be interested in getting more users to visit the site and load multiple pages, or getting core users to return regularly, or to tweet about what they saw to all of their followers, or to buy something from a sponsor. These various scenarios fall into different categories and require different approaches.

In my next post, I’ll address a simple framework for the different types of metrics that most modern web destinations should be paying attention to, including a few that may surprise you.