Arijit Sengupta, CEO of San Mateo, Calif.-based BeyondCore thinks a mismatch has developed between marketers’ needs for complex data insights and the existing analytics resources available to provide them, he told CMSWire in a recent interview.
That gap between problem and solution has left today’s CMOs in search of “Goldilocks” scenarios ―not too superficial, not too technical but just right ― while they wait for next-generation analytics applications that will be up to the task of extracting meaningful data insights while remaining intuitive, visual and easy to use.
Three Bearish Problems to Avoid
Sengupta pinpointed three issues that he thinks are currently standing in the way of better alignment between enterprise marketers’ questions and the analytics they need to answer them.
The most obvious ― and most intractable ― is that enterprise demand for data scientists with the skill sets to supply marketers with actionable insights from complex data sets continues to outpace supply.
A more subtle factor according to Sengupta is that traditional analytics applications can be intimidating to marketers because they often more closely resemble coding environments than user-friendly, visual dashboards.
Yet a third problem, which impacts programmers and marketers equally but manifests differently in each case, is the potential for logical bias to creep in, leading to the wrong conclusions based on the wrong data.
Bias Is Everywhere
In the case of programmers, Sengupta pointed out that the custom-configured solutions produced by so-called Integrated Development Environments (IDEs) are only as good as the coding assumptions that underlie them.
The risk is that although marketers may assume such analytics solutions to be highly trustworthy, they are actually subject to inadvertent coder error and bias.
Yet Sengupta points out that analytics solutions that get implemented at the business unit level while bypassing IT, introduce equal but opposite risks.
“Self-service analytics applications [that] empower marketers to draw the wrong graphs ― and therefore faulty or premature conclusions ― based on insufficient data mean that business leaders could be making decisions based on inaccurate or misleading information,” he warned.
CMOs Are Worried
Sengupta is by no means an outlier in his assessment that traditional analytics are leaving today’s marketers high and dry.
In research based on a survey of 524 CMOs, IBM’s 2014 Global C-Suite Study indicated that 82 percent of CMOs felt underprepared to deal with the data explosion engulfing them, up from 71 percent who expressed that feeling in 2011.
Deloitte made similar findings in its recent CMO-mentum report which revealed deep marketer concern regarding the gap between their organizations’ expectations on the analytics front and the results they have the expertise to achieve. While 66 percent of the CMOs in the Deloitte study relied on analytics to make key decisions, 71 percent admitted that they were not realizing the full potential of their data.
Mindset Is What Matters
Where does this disconnect leave marketers who are searching for their Goldilocks solutions? If self-service business intelligence is no longer enough and enterprise data science isn’t available, what does Sengupta suggest?
He believes that marketers should cultivate the mindset that “what [really] matters is asking the right questions of your data. True analysis involves finding the right graph, making sure the pattern is actionable and properly understanding why something is happening, not just what is happening," Sengupta clarified.
Innovations Around the Corner
Sengupta looks to three on-deck innovations in the analytics marketplace to help marketers close the gap between good intentions and quality execution. He feels that:
- Automated smart pattern discovery will help marketing decision makers correlate multiple data sets to reveal hidden patterns of business intelligence, while rendering ask-and-answer analytics approaches obsolete.
- Assisted Interpretation will employ statistical analysis on behalf of marketers to aid them in determining whether the patterns they detect are merely visually appealing artifacts or bona fide insights that are accurate and reliable.
- Prescriptive recommendations will elevate decision making beyond simply identifying which external factors can be influenced and into the realm of specifying the best ways to do so.