My linkedin answer adjudged as best!!

A while ago, there was an interesting question posted on linked Q&A. Had written down my thoughts and it seems the person posting the question was happy with it. Nevertheless, would like to share my thoughts with the larger analytics community and look for comments. So here goes..

What are some examples of using "alternate talent" for knowledge intensive industries like Research and Analytics? In addition, what are best practices in training and development of such talent?

Great question pertinent to the current situation. I will answer the questions using analytics as a context as that is where I have come across and attempted addressing the problem.
First is to accept the market reality that unlike in the US, in India, the analytics professional is not the the superman or superwoman that you see elsewhere (US or Europe) - meaning he/she will not have a above average understanding of statistical concepts, great working knowledge of tools like SAS and more than sufficient business understanding to make the solutions relevant. At best, we can find such resources at a slightly senior level (4+ years into being in this industry). The entry streams are a set of distinct capabilities - A statistician or a SAS programmer or a business consultant (recent MBA grad).
Second is to determine the set of capabilities that is needed to answer to a certain level of need - what level of "statistical soundness", "analytical orientation" and "business relevance thinking" that is required to address the expected level of need. For example, you might say, the variety of problems that come by your way only require x amount of statistical thinking, y amount of business thinking. Largely we have seen problems can be decomposed into the repetitive part and the customized part.
Third is hire right i.e to staff the team based on the capability analysis you have done (this is where the alternate talent comes in). This means if I am starting an analytics unit, I should not be looking at staffing all my modeler requirements with IIT/ISI graduates or staff all my project management/ client interaction with MBA graduates from premier schools. Instead staff the team with a couple of folks who are displaying sound statistical thinking (on a 1: 10 basis if you have a fairly clear visibility on the kinds of projects) and a couple of folks from the premier schools for bringing in the business relevance/ marketing thinking into the solutions. These people will be performing their regular roles of delivery and will also put on their statistical/ business hats when required. The rest of the team members can be bright engineering/ arts/ science graduates who are capable of getting themselves equipped with the right skills given the motivation and infrastructure. I have recruited from research institutions/ think tanks and engineering colleges and seen that the enthusiasm that these people bring to the team are no less than what you will see from those coming from mainstream learning areas (MBA. MSc statistics).
Fourth is to invest in training. This can mean either having a tieup with external organizations that can train the employees into becoming analytics ready, or have an internal training system developed over time customized to the organization. Tying up the training to appraisal ensures that it is taken seriously. Something that has worked for me is making people go through evaluation on different tools/ concepts/ techniques and make them aware of their need to do well to progress. Fifth is to invest in knowledge management and bench. Last and definitely not the least is that when you are in this role, you are constantly on the look out for people who have a passion for analytics, strong customer/ client orientation and a high level of ownership. Referrals are also a good resource. The traditional recruitment consultant route will not work here as they have not evolved well to understand the needs of the industry.

When numbers speak for themselves

"The problem is not ignorance. It is preconceived ideas" says Dr. Hans Rosling. Dr. Rosling's speech in the annual TED forum debunks a lot of myths about the developing world using data. The animation is fantastic and so is the narration which we can see comes with a lot of passion. An excellent piece of material that can be used to show how data analysis can provide wonderful insights.

Incidentally the TED talks are something I recently stumbled on to and they definitely seem worth listening to once a while.