Forecast 8 to 40 Inches

If you forecast the next quarter’s sales to be between 60% and 400% of last quarter, you would probably very quickly be shown the door. But that is what we have here right now – the local forecasters are saying that tomorrow’s snowfall is expected to bring between 8 and 40 inches. With the National Weather Services $900M budget in 2012, I expect them to be a little more accurate on these things. When dealing with your financial matters as well or other factors in the marketplace, error bars are critical to predicting numbers that you can act on. Whether ordering more parts for products, scaling back resources, or building out a new geography, having a reasonable amount of certainty in the outlook is paramount to running a smooth (and profitable) operation.

So how does the Weather Service have such uncertainty in something happing a day away, and how can you have a much better view on your future prospects? There are some similarities between weather and business as well as some differences.

• Historical data. Wonderful data abounds about every aspect of weather for the last hundred years or so. Similarly your business often has a track record that is well known. But taking this data and extrapolating means that you have not only an understanding of when sales jumped or slumped, more importantly you know why this happened. It’s not just about the numbers, though the naïve can always apply a small uptick to last year and call it a growth forecast. Knowing the important factors that influence your business, such as an El Niño or La Niña effect, is what determines whether your forecast is reliable.

• Nice Models (No not Gisele or Lars). The weather folks have spent lots of time and money creating beautiful models that predict events such as tomorrow’s storm. In the words of the NOAA, “Most people don’t realize it, but weather models are one of the main tools forecasters now use to help predict the weather.” They have their storm track, precipitation, upper air, pressure, etc. all spinning around in their supercomputers to tell us that we will have snow tomorrow. I think the “red sky at morning” is similarly accurate sometimes. Combining the data and effects together, you often can create a reasonable predictive model for future events. This is often done in a more informal fashion, but is a way of increasing the confidence of your forecast since it considers many factors at once.

• 20/20 Hindsight. Of course today you can perfectly understand why last quarter (or year) was such a boom (or bust), but applying that to the future is important. Many times, you are into the next quarter and all of last quarter’s assumptions are just forgotten. You need to dredge up your forecast AND the rationale, and compare this with your new hindsight wisdom. This will help you to learn which assumptions held up, which were blown up, and what new affects you did not see coming. Then apply this new-found perspective to the creation of your future views.

• Good Snowplow Drivers. With all the wonderful forecasts guiding your operation, execution is the most important part. All is usually fine until the storm veers off and freezing rain starts to coat the roads. Having the ability to adapt to changes in the marketplace, competition, and your own organization’s foibles is very valuable. This can often make for smooth driving even when the weather does not cooperate.

So whether we get will 8 or 40 inches tomorrow is a great mystery at this time (though it may be obvious by the time you read this). Though the Weather Service has a rather checkered record of longer term forecasts, and a slightly better one for short term, you need not be held to this rather low standard. Joining a realistic understanding of why you had certain performance in the past with a solid grasp of the marketplace and its customers, you can craft a confident forecast of what will transpire.

PS. I’m hoping for the 40″