One of the biggest problems with many American companies is the loss of nimbleness. Corporations are so weighted by forecasts and metrics they no longer can react with agility. I believe too many business decisions are based on looking through the rear-view mirror. 

There are two serious problems with using past data to predict future results. First, historical performance, unless repeatable at will, is no indication of future results. Every investment must caution the potential investor of that very fact. You’ve heard those words many times: Past performance is no guarantee of future results. Then why do we base nearly all business decisions on past tense data?

We forecast this year by imposing an adjustment, a guess at best, on last year. That’s not a forecast, it’s little more than a guess. Sure, the guess may have some statistical support which could possibly put it within a reasonable range if everything goes accordingly. But it seldom works that way in the real world.

As we look at the past, we tend to smooth it. We ignore extreme data points that distort the averages. Sounds like a reasonable thing to do but it’s not. It’s not reasonable because we will always have extreme data points. The extremes have the most impact on operations and are the hardest events to predict. Yet we ignore those extremes when we smooth data by averaging as our first step in forecasting.

Granted, we cannot predict what extreme event might occur. Perhaps it comes in the form of a huge order pushing a particular month’s sales up by 500% over plan. The influx of business and pressure to meet on time delivery cause chaos. Plans unravel. Tensions rise. Costs sky rocket. What should be a welcome event becomes a crunch and perhaps a loss.

The extreme event could come in the form of a rapid business down turn. Then what? The business has commitments to buy materials based on forecasts. Inventory piles up. Margins shrink. The cost cutting begins.

What can be done? Surely a business has to forecast. Yes and no. Yes, a business needs a best guess at what, within reasonable tolerances, will likely happen. But this guess is a component of a plan. And the plan should include some steps to deal with extreme events should they occur.

This where the lab comes in. I think it would be a good practice to discuss extreme scenarios at business meetings. What if sales landed a huge order? Who would do what, when to maintain a smooth flow of operations? Does the business have access to backup suppliers to be sure the material needs are met? Can labor be scheduled to meet ship dates? These questions need to be asked and then experiments run in your lab.

Run the scenario. Call your suppliers and ask the questions. If you needed 5,000 pieces next month instead of your usual 500, could your supplier get them to you? Ask. Calculate the labor and make a schedule. Can it be done?

Who has time for this? You do. Running the lab exercise is very beneficial to your business and could potentially make it far more profitable. In some cases, having run a lab or two could save a business from failure when the unpredictable happens.

What is incredibly ironic is that the unpredictable, is very predictable. Extreme events WILL happen.

In science, we conduct experiments to prove hypotheses. The experiment should give us an ability to predict results. I know from doing the lab work that water boils at 212 degrees F. I know from experimentation that adding salt to the water causes it to boil at a lower temperature. If I need to boil water faster, I can add some salt to it. This approach needs to be applied at your business.

Okay, this sounds like a good idea but really, what’s the urgency? Remember, I started by saying there are two reasons rear-view forecasting is a bad practice. Since the past does not predict the future, what’s the other reason?

Simple. Object in mirror is closer than it appears. Those extreme events will happen sooner than you expect— every time.

What I want you to do is spend less time measuring and analyzing and spend more time in real simulation work. Lab work.

Chris Reich