Most companies heard about value of data-driven decisions, and every company accumulated more data than they know what to do with. Some enterprises invested in technology tools sold to capture transactions and mine oceans of data in hopes of increasing their revenues, reducing their expenses and improving every other KPI in between. While dashboards and interactive reports present us with an abundance of analytics results in real-time, I would like to ask you Dr. Phil’s question – “How’s That Working for Ya?”
For many people. distilling insights from analytics is challenging enough—translating those into specific actions is often impossible.
“A perfection of means, and confusion of aims, seems to be our main problem.” Albert Einstein
The good news is that you cannot buy another tool or technology to resolve this problem for you. If it was, our unemployment rate would be even higher. The job of setting right aims is still best performed by humans. Setting of aims should not be confused with creating hypotheses or forming assumptions. If you do that, you limit potential actions only to validation or rejection of those hypotheses. Meanwhile, the most valuable potential actions may never discovered.
I’ve seen the best results when:
1. The scope is focused on 3 to 5 potential targets directly associated with specific business goals at the time:
a. “Is there an opportunity to disrupt the XXX market segment?” or
b. “Is it possible to improve our customers’ experience by linking it to a performance and compensation model of employees from departments other than sales and support?”, or
c. “What is the confidence of executive management in our sales forecasting process?”
2. The targets are potentially achievable, i.e. specific actions could be taken if the insights, guiding such actions, were known. In other words, you have authority or influence to act upon insights to reach the selected targets. If you don’t, the best insights are not likely to result in decisive action that requires changing somebody’s mind when it is already made. The insights that the bourgeoning MP3 player market segment could be disrupted by redesigning user interface and offering a source for high quality digital content, would not help to introduce iPod revolution without the influence of somebody who had an authority to act on such an insight.
3. Appropriate data sources can be identified and triangulated. The intersection of signals from multiple sources provides better guidance for action. Think of it as GPS requirement for signals: a minimum of 3 satellites are required to produce relatively accurate position information. The minimum accuracy (confidence level) and volume (statistical representation) factors have to be assessed before the aim is set. There is no point in producing insight that does not give you confidence to act.
The short-term, “ready, fire, aim” approach is not likely to produce meaningful ROI from a technology investment into analytics.
This is one of the major issues that is the result of big data. We have the data- but do we know how to properly use it? Data itself is meaningless. The only way to boost the bottom line is to implement changes that are backed up by this data, which requires complete and thorough analysis.