CASE INCIDENT 2 Big Data for Dummies Do you need big data? Maybe the question is better quick answer seems to be to hire talent. But not just anyphrased as: Can you afford not to use big data? The age of one will do. Here are some points to ponder when hiring big data is here, and to ignore its benefits is to run the risk data professionals: of missed opportunities. Organizations using big data are quickly reaping re-1. Look for candidates with a strong educational back wards, as a survey of 2,022 managers worldwide indicated ground in analytics/statistics. You want someone why recently. In fact, 71 percent of respondents agreed that . knows more than you do about handling copio. organizations using big data will gain a "huge competi- 2. The ideal candidates will have specific experience it data: 58 percent responded that they never, rarely, or only your industry or a related industry. "When you have sometimes have enough data to make key business deci- those Ph.D.s in a room, magic doesn't necessarily hat sions. Furthermore, they've witnessed the benefits: 67 per pen because they may not have the business capabilit centagreedthatbigdatahashelpedtheirorganizationtoinnovate.Sowhydidonly28percentfindthattheiraccesssaidAndyRusnak,aseniorexecutiveatErnst&Younh3.Searchforpotentialcandidatesfromindustrylead innovate: So why did only 28 percent find that their access to useful data significantly increased in a year? organizations that are more advanced in big data. Aecording to Amy Braverman, a principal statistician 4. Communication skills are a must. Look for a candidate who analyzes NASA's spacecraft data, the problem is in "who can translate Ph.D. to English," says SAP Chici interpreting the new kinds and volumes of data we are Data Scientist David Cinsberg. He adds, "Those are, able to collect. "This opportunistic data collection is leadthe hardest people to find." ing to entirely new kinds of data that aren't well suited to 5. Find candidates with a proven record of finding usethe existing statistical and data-mining methodologies," ful information from a mess of data, including data she said. I and business leaders agree: in a recent sur- from questionable sourses. You want someone who is vey, "determining how to get value" was identified as the analytical artd discerning: number 1 challenge of big data, 6. Look for people who can think in 8 - to 10-week With strong need combating the high hurdle for usabilperiods, not just long term. Most data projects have a ity, how should a company get started using big data? The short-term focus. 7. Test camdidates' expertise on real problems. Netflix's increase profits. What data might be available Director of Agorithms asks candidates, "You have increase profies. What data might be available this data that comes from our users. How can you use would be important to your decision? it to solve this particular problem?" A1-19. What kinds of data might we want in OB Queatlone applications? 1-18. Let's say you work in a metropolitan city for a large 1-20. As Braveranan notes, one problem with big data department store chain and your manager puts is making sense of the information. How might you in charge of a team to find out whether keepa better understanding of psychology help you ing the store open an hour longer each day would sift through all this data