Five Steps For Making Data-Driven Decisions

Piyanka Jain was head of business analytics at PayPal North America until last year and now runs Aryng, which conducts analytics training. Aryng teaches a five-step process that it calls the Data to Decisions framework, which Jain outlined for me over the phone:

1. Understand what the real business question is. Ask the stakeholder what she really wants to know. “Why is this person asking this question? What’s the context, what are the impacted segments?”
2. Create an analysis plan with hypotheses. “What are your fundamental beliefs about this problem? Let’s do a guided exploration. This is where you decide what methodology you’re going to use–am I going to use correlation analysis, is this a profiling problem, am I going to go into predictive analysis?”
3. Collect data. Based on step 2, you’ll know what sort of data you need to collect.
4. Gather insights. “Now that you’ve collected data, you’ve ordered it, validated it, triangulated it, then you get into the analysis part based on the technique you’ve chosen.”
5. Make recommendations. “It’s not only the technical aspects that are important, but also the soft aspects: what are you doing with your stakeholders? You are an analyst. What are you doing with the folks on the other side? How are you building alignment so that when you say ‘here are my insights, ta-dah!’ somebody is going to be able to work along with you. If you’ve not brought your stakeholders along, then they won’t necessarily take your recommendations.”

Jain now blogs for Forbes; follow her here. An informative white paper about her work with a winery is available on Aryng’s Web site.

Jon Bruner, Forbes Staff
Forbes.com

Original article found here- http://www.forbes.com/sites/jonbruner/2012/04/20/five-steps-for-making-data-driven-decisions/