Next level analytics competency – part one

The presentation addressed the main issues in the shift from data as an IT asset, to data as a main asset for decision making.
I took a look at traditional business intelligence ecosystems and how we all thought it was going to be done: the perfect situation of taking data from source systems, creating a unified data model and data store and thereafter, having a set of business users accessing only that data model.

The reality is that business has never really been happy with this approach. This approach can cause a lot of red tape, and rightfully so. Incorporating a new source system into a data warehouse, requires a lot of planning, data analysis, identifying business rules, cleaning data and ultimately, making it available to business. The red tape almost makes me happy – it means that data warehouses or business intelligence environments are mission-critical systems, changing them shouldn’t be an easy process and they should be treated as core business systems. The by-product of the red tape however, is that now, organisations typically have several rogue MIS systems that IT departments often have no knowledge about. Sometimes, these systems are running on local machines and have no backup or business continuity strategies. This isn’t a bad thing by any means and if anything, it shows that people are willing to break the rules to ensure they can make data-driven decisions.

So what are the steps to creating a data driven culture? Eric Schmidt of Google says, “We run the company by questions and not answers,” and this speaks to the first step in the process, which is to to define the question. We then need to choose the right data to get what we are looking for. Big Data is about finding an answer to a question that we never knew we had – pick the right subset of data and use just that. Don’t bloat your initiatives with too much data from too many data sources. For example, if you don’t need users’ tweets to get to the answer you are looking for, then don’t include them in the data set you are analysing. The last three steps in the process are self-explanatory and are repeated. We need to analyse the data, present it through an intuitive interface and make the decision.

In the second part of my presentation, I looked at different ways in which organisations can develop analytical talent as well as how to transform culture, capabilities and technology.

Related Articles

Customer data privacy:
Trust or trapped
MVP vs big bang release
Effective prioritisation:
Not all requirements are created equal