Friday, September 14, 2018

Understanding and Using Data: Waffle House Example

Comments on, Why FEMA is Monitoring Waffle House this Weekend

look outside your organization for risk indicators that might help you make better (and faster) decisions, particularly when those risks are activated. Second, that you should explore crowdsourced risk data as a source of up-to-date information.

I agree. Also, just look at data close to where the action is (internal or external).

And don't just look at aggregated data but dig into what individual data points tell you. Aggregated data is very useful but it also can mask meaningful insights available when data is looked at more closely. It isn't a perfect match to Waffle House data but I think the principle is visible in the Waffle House example.

The Waffle House closure data is based on actually closing based existing conditions while warnings and evacuation recommendations are based on predictions about the weather and the impacts those will have on locations. The warnings are necessarily predictions (to be useful for the whole community they need lead times to take action) where the Waffle House has more flexibility and the organization has managed their system to be more capable of adapting to harsh conditions. There is a real similarity with designing a agile software development process that is able to be more flexible and react quicker than old "waterfall" style organizations that have to predict far in advance and adapt slowly as conditions change.

Waffle House closures are a useful data point for conditions on the ground that FEMA can use with other data to make decisions on how to react.

Related: Bad Weather is Part of the Transportation System - Leadership: Taking Action When Others Are Unsure - Lessons on Competition from Mother Nature - All Data is Wrong, Some is Useful