1 Analytics Candidates as just SAS/ SQL experts
Very often we see analytics recruiters grilling candidates on just SAS/ SQL skills. These might be relevant at roles that require SAS/SQL as primary skills and not really modeling/ number crunching to a large extent.
For eg: back end teams that convert models into canned SAS/ SQL codes for reuse. Else, its completely wrong to see analytics candidates as merely SAS/ SQL experts. The problem exemplifies in cases of outsourced recruitment to job consultancies, who have a tradition eye towards Indian IT i.e. all functions seen as mere skills into IT tools.
2 Analytics Candidates as just Statistics Experts
While its required by even a fresher aspiring to get into analytics to have a working knowledge of topics like regression, isn’t it too demanding to grill them on say “Agglomerative Hierarchical clustering with cosine similarity”. And we quite often see this happen in the industry. While it is good to test candidates on basic and sometimes even advanced statistics concepts, analytics is much more than that. And for all your needs of deeper analytics skillsets, a quants PhD is someone you should be looking at (which brings me to the next point).
3 Quants PhD’s as Analytics professionals
A PhD in quants is a great degree to have. It is highly paid and respected in the industry. It gives a recruiter confidence on a candidate’s ability to dig deeper into a topic and research to its core. But the buck might stop there. People with a strong statistics orientation may tend to focus too much on deriving elegant mathematical solutions, where as all the business may need is a viable solution. Very often we see businesses trash excellent quantitative models because they find them of no use to day-to-day business. Having said that, if you find a quants PhD with deep business acumen and data skills, consider yourself lucky and be ready to put in big bucks.
4 No Emphasis on Business Acumen
Most of the times we see recruiters not emphasizing enough on business acumen. Analytics recruitment is tricky. The best results are attained by those who see it as a well-rounded combination of Data/ SAS skills, Statistics knowledge and business acumen. Try to question why LTV rather than how LTV or why clustering rather than How. “How” is easier to learn, “Why” requires much more imagination. Grill candidates on how analytics can be used for promotional campaigns, reducing portfolio risk or personalization.
5 Don’t get stuck on relevant years of experience
We see candidates jumping into analytics at every stage of their career. Its good to see how much relevant experience the candidate brings in to his current role, most often recruiters completely disregard previous experience into unrelated field. While this might be pertinent in some situations, its worthwhile to give some credence to earlier non-analytics experience as well. This also brings us to a growing issue in analytics manpower currently i.e. hordes of mid managers in analytics with little relevant experience. Its appropriate in these situations to place candidates in roles that helps them grow significantly in lesser time than their peers.