Nowadays we evangelize that businesses should be data driven. However what we are not speaking about is that by saying data driven we do not mean taking all numbers literally but reading them as a trend. If we put things in a context the objective of web analytics data is to measure the performance of your marketing activities either being your website or your PPC campaign etc’. As such what web analytics is good in doing is identifying you the good and the bad performers on your website and provide you actionable insights provoking you to take actions to optimise your business. Although with web analytics we can measure marketing ROI is not the practice which will help you make financial decisions on your business profitability as a whole as it is out of scope.
This is one of the limitations that we reach due to the medium we are measuring as:
- The web although is a control medium nevertheless the web users are so fractional and random that complicate the process of measuring and tracking.
- The web analytics tools we are using have technological limitations that does not let us know important information i.e. we do not know how much time a visitor spent to the last page before exiting.
- The web is spreading across many platforms: PC, mobile, iPads etc’ giving a further layer of technological complexity.
Therefore when we evangelizing a data driven culture we want also to set a clear framework:
- Objective: Measure marketing activities.
- Limitations: What we cannot do and what are the ways around it.
- Data literacy: We want to explain how the data should be read to make sense and provide insights
In a Data Driven culture we need to create data literacy at list to the stakeholders involved and the report recipients so that everyone has a basic understanding on how the data should be interpreted.
Numbers are measurements of a given context for a specific objective, serving a number with ought a context is like serving food with ought its plate. This is a very important aspect and we should always clearly communicate what a number represents by giving at a minimum the following context:
- Objective: What is the purpose of this number/formula what do we know if it is high or low? Can we know an average value? (i.e. pageviews - objective: to see how many people see the pages on our website)
- Calculation: How the number was calculated? (i.e. pageviews: were taken form our analytics tool which calculates 1 pageview every time a visitor sees a unique URL)
- Data errors: What are the know/expected cases that lead to data distortion and in what percent? (i.e. pageviews: a page is calculated every time on refresh, the same page is calculated more times if you have different pages with same URL etc’)
Until now web metrics have been this thing that resides only on the analyst mind. My belief is that by educating and bringing clarity to as many stakeholders in our organizations just helps be everyone more productive efficient and happy 
