My friend and colleague, Mark Weiner, CEO of PRIME Research, LP, recently wrote “Driving the Action: The Ascent of Insights and the Decline of Data” for PRSA’s The Public Relations Strategist. His excellent article closes with four guidelines for transforming data into insights:

1. Monitor and integrate. Good data is essential. Data may come from a variety of sources including social and traditional media, blogs and survey results, along with business measures such as employee loyalty, customer retention and sales volume or, as in MasterCard’s case, transactions. These “small data” streams lead to “big data” insights when combined by expert data scientists.

2. Explore and investigate. The speed, volume and variety of conversations, topics and issues at a given time mean that everyone must pay attention to the thousands of mature and emerging issues bubbling around us.

3. Analyze and illuminate. Relying on talents, skills and tools, data scientists must collaborate with PR leadership to uncover gaps, unusual activity and other signs that require further inquiry. Reduce the many uncovered opportunities and instead focus on those with the greatest importance to the organization, and with the greatest opportunity for positive differentiation.

4. Share and initiate. Insights without action are the slowest path to victory. Since meaningful insights are rare, share them with peers to optimize your “ROI (return on insight)” and act on what you learn.

I thank Mark for his great article and guidelines. But I’d like to add a critical piece I feel is hinted at but not explicitly stated in his guidelines.

In all the research work that I have done, the piece that is critical to turning data or information into insight is a deep understanding of the goals and strategies of the client organization for which you are doing the research. In the MasterCard example Mark refers to, the insight was that there was a business problem, not a communications problem. That required understanding MasterCard’s business objectives:

… Those who had already tried the [mobile payment] technology expressed two sources of anxiety: how readily merchants would accept the payment method and the level of security. An even deeper statistical analysis showed that unease over acceptance had come about from certain vendors — taxi drivers and fast food restaurants primarily — and their inability to process mobile payments at the point of sale.

Extracting meaning or insight from data is what good research is about. Sometimes we design the research and ask respondents questions. In the case of large data sets, which is what Mark writes about and which is what we most often deal with when dealing with social media analysis and big data, we can query the data.

One question Mark’s team asked the data was “do social media commenters like the payment method concept?” Then, one of his team members was clever enough to ask “is there a difference between how much people who have used the payment method like it and how much people who have not used it like it?” The answer to this question was yes. What was interesting was that triers were less excited by the new technology than non-triers. The obvious next question is: Why?

So this line of questions turned out to be golden. It showed a potential problem with the offering. And while these questions are simple and straight-forward, they are directly linked to the business success of a product. MasterCard is looking for repeated use, not just trial. The analyst who asked the second question knew that.

So, regardless of whether you are designing a survey, querying big data or swimming through it looking for “insights,” I would argue you have no hope of discovering meaningful insights if you do not have a solid understanding of your client organization’s business goals and strategies. And this is as true for communications practitioners as it is for data scientists.