Operations Research from the Master

Our feature on Proctor & Gamble's use of operations research â€" a burgeoning discipline that utilizes deep mathematical analytics to drive all sorts of business decisions â€" is getting some attention, and from the right people. P & G's own OR guru, Glenn Wegryn, has weighed-in in the comments section, answering reader questions and concerns.

When readers "chf" and "rich" jotted down their concerns about the limits of OR effectiveness in a topsy-turvy world, Wegryn - who goes by the handle "Analytics4Me" - stepped in and offered up a tidy alternative definition. He said:

It's possible that you may be defining "OR" or "Analytics" a bit too narrowly -it is much more than linear programming methods, which I completely agree have limited scope if you use them in purely textbook (non-industrial strength) problems.

Our shop utilizes a very robust suite of tools â€" linear, non-linear, decision analysis, dynamic simulation, Monte Carlo simulation, fundamental and advanced statistical analyses, etc. We believe strongly that a good analytics practitioner can't have too many quality tools in their tool belts, because at the end of the day you are helping a decision-maker make the best decision possible in light of quantitative and qualitative constraints and uncertainties. It's the role of a seasoned analytics professional to choose the right tool for the job â€" and THAT's what makes this work so fascinating!

How does a small or midsize company take advantage of all these slick OR tools? Or is OR by definition a practice that ideally requires big scale?

Small and medium-size companies can leverage several external sources to get started, such as:

- Analytic consultancies, specializing in end-to-end analysis
- Larger software vendors who have an on-staff consulting capability (while, of course they are attempting to sell their software, piloting for a reasonable fee lets you try before buying and better understand the capabilities the analytic software can bring)
- A number of universities with quant departments are always looking for businesses to consult with and give their grad students real-life experience on business problem solving.

As a business develops a better understanding of analytics capability, it may warrant investing in an on-staff analytics resource â€" preferably someone who is already within the company who understands the operations and culture of the company. The resource's time could be split between in-house analysis; developing improved data streams for analysis, and coordinating additional analytical work from any of the above sources.

See also our Intercom follow-up on OR resources.