Lean Analytics with Tapstream

A pivotal process of the product development cycle is measurement, evaluating the results of the labor and effoiturt on a product, validating one’s assumptions and assertions. Analytics allows you to track the metrics that are most crucial to evaluating the performance of your business, an on-going testament as to how effective your customer acquisition strategies are, essentially your overall customer traction. The first part of this article will articulate the importance of lean analytics, with the second half showing you how you can utilize Tapstreams fantastic suite of tools to measure just that.

What is lean analytics?

The question is, how do you know what metrics you should track, that will be a true reflection of how you are progressing?

Lean Analytics by Alistair Croll and Benjamin Yoskovitz (2013) places emphasis on useful metrics being those that are comparative, such as tracking increase conversion from one period to another. That is, a metric that is measurable and understandable, which generally involves a ratio or rate, and whereas many startups fall into the trap of vanity metrics, statistics that whilst can make you feel good, are not actionable metrics.

Total Signups is an example of a vanity metric, a value that increases over time but does not provide insight into how active users are. Total Active Users according to Croll & Yoskovitz, is a slightly more insightful, but similarly will also produce an ever-increasing gradient graph, and thus is also vanity. What is the real actionable metric is the ‘percentage of users who are active’, a period-by-period comparison of how you are doing, as opposed to an always increasing metric. So anything that measures something over a specific period of time, based on marketing initiatives is actionable and good. However, that introduces another problem, establishing causality.

That is, when you are measuring something, you have to establish whether the feature or marketing effort you just completed is the cause for the metric change. This is causal establishment, and whilst this is a big topic outside the scope of this article, let’s just say that you need to find a way to correlate a feature or initiative with a metric, not as a result of another cause. Are more users becoming active because of a twitter campaign you just concluded, or because of a feature you just completed?

Read my full article at: http://www.programmableweb.com/news/lean-analytics-tapstream/analysis/2014/10/28-2