How often should you sense demand?

When I posted about Agile Demand Planning a couple of days back, some of you wanted to know the mechanics of creating this agility. Let me tackle the first part of it in this post… how often should you sense demand?

Let’s check some other sensing mechanisms we experience in our daily life.

How often should Google Maps sense traffic on various routes and update its estimates? Would you be happy with once a day updates? Hourly? No, we are used to continuous updates and we would want nothing less. If there is a vehicle breakdown somewhere, we want to know about it and avoid that route.

Let’s take another example. We have temperature sensors in various appliances like refrigerators, air conditioners and geysers. These sensors feed signals to thermostats for either cutting off or restarting the operations. Would you be happy if these sensors operated only once every hour? No, we want them to sense continuously.

Why should it be any different for Demand Sensing? Systems Dynamics and Servo Control theories tell us that any system operates smoothly only if the output sensing is continuous and a feedback loop exists to control the input accordingly.

To get best predictions, frequency of Demand Sensing should match the frequency of demand data available to us.

Do you still want to depend solely on your monthly forecasts?