A shallow talent pool of skilled workers to analyze big data, combined with the challenge of weeding out bad information, continues to cause nightmares for CIOs.
Khalid Khan, partner at A.T. Kearney.
Two-thirds of companies that possess even the most advanced analytics capabilities cannot hire enough people who can generate insights from corporate data, according to new research from A.T. Kearney, which surveyed 430 senior executives. Moreover, companies will need 33 percent more big data talent over the next five years, says Khalid Khan, A.T. Kearney partner and co-author of the research.
The stakes are high for attracting and retaining analytics talent. As companies digitize their businesses, it's imperative to mine the resulting data for correlations that help refine products, make processes more efficient or even identify new revenue streams. For example, the Weather Company is mulling how to predict weather patterns to help retailers divine whether or not they need to increase or reduce their inventory of water. But while business cases for big data appear infinite, the talent pool to mine it remains small.
In search of triple threats
Khan says companies should look for "trilinguals," or people with a firm grasp of quantitative analytics, digital technology and business strategy. He says respondents who qualify as analytics leaders -- those who have baked analytics into the decision-making process to predict trends that will shape the business and drive competitive advantage -- have been the most effective incultivating trilinguals who possess both business and data skills.
Few such analytics hybrids are available in the corporate sector, forcing many companies to court college graduates who are technically proficient and schooled in statistical modeling. But finding qualified candidates this way is rare. They often possess the right technical analytics chops but lack the ability to derive business insights with the data. Nearly 60 percent of respondents say that the talent coming out of universities is insufficiently prepared.
Seeking to meet companies in the middle, schools such as Carnegie Mellon University, North Carolina State University and New York University have built cross-disciplinary programs to help train data scientists, Khan says. But they aren't producing candidates fast enough given today's talent requirements.
If there is a silver lining for companies that are having a hard time cultivating analytics talent, it's that most organizations are still struggling with rendering their data for analysis. It's a case of technology trumping organizational readiness. Analytics potential has grown sharply, sparked by the emergence of tools such as Hadoop and Spark, which process large amounts of data at a lower cost than was previously possible. To prepare data for processing, companies are extracting it from many siloed applications and systems. They must then normalize it, essentially weeding out bad data that skews results. It's a nontrival, time-consuming task, as Merck Animal Health CIO Dave Williams recently told CIO.com.
Sign up for MIS Asia eNewsletters.