What If...You could blunt risk?

By Sean Hartnett, CSP

Just as there are statistically no two identical snowflakes, there is no winter season that is statistically identical to another. The challenges retailers, suppliers and contractors face are the same year over year: “Should I expect this winter to be better or worse than last season?” It is incredibly easy to be caught in the speculative trap considering the past four years have had above-normal snowfall nationally.

These facts can be misleading to the untrained eye, causing many organizations to build business plans based on emotions rather than a more practical solution. It is important to recognize that weather, especially snow events, statistically repeat less than 20% year over year in any individual market. Therefore, if you plan your snow business decisions based on last year’s weather, you are potentially at risk of missing the mark 80% of the time.

Given the recent harsh winters, contractors may fall into the trap of building a strategy based on emotion rather than on statistical data.

To steer clear of the speculative trap and take a more practical approach for next winter, I recommend utilizing a statistical approach. There are three primary methods available to the market:

Averages
This method compiles an annual snowfall sum data set for an extended number of years (5, 10, etc.). The average of snowfall values is then used as your baseline for developing a plan. Applying averages as your baseline is a simplistic and practical method for preseason planning. In statistics, an average is a measure of the “middle” value of the data set (snowfall). The average is a single value that is meant to symbolize a list of values. The benefit of using an average is that it provides a baseline of the central trend of weather over time.

Although averages are a step in the right direction for practical planning and understanding snowfall trends, they do not provide full visibility or tell the whole truth. Averages only explain the central trend of historical snowfall data sets. This presents a problem because annual snowfall almost always has a skewed distribution of the values with some small numbers of very high values. In the example below, using averages from Baltimore, MD could be problematic given the 2009-10 winter was such an anomaly.

Weather-neutral planning
The next level to utilizing a more practical and actionable plan requires organizations to seek assistance from the outside, such as a business weather intelligence firm. At Planalytics, we employ Deweatherization®, which is the method of objectively and statistically removing weather’s volatility from historical results. The value of this method allows organizations to plan from a “weather-neutral” position. The outcome is a more accurate plan built upon the statistical reality that weather events one year to the next will trend directionally back to the “norm.”

Weather-neutral planning is idea for preseason planning decisions such as operations, budgeting, procurement and sales. The example above represents planning for bulk salt purchasing.

Recommended usages for this method are demand planning/buying, season timing, allocation/replenishment and retrospective reporting. For example, a contractor in Pittsburgh who is in the process of preseason planning is faced with a decision about how much bulk salt to purchase. The contractor in this scenario applied a Deweatherized® value as his baseline, which enabled him to have a more accurate plan for the season.

If we look ahead to this coming season, nationally, this winter should be more like a typical La Niña winter. A La Niña winter produces a northern snowfall track without the extremely strong blocking North Atlantic pattern. This would result in a decrease from last year as systems move through more quickly and produce less snowfall.

To translate this to snow removal, the example below highlights the Deweatherized® outlook for the Northwest region of the U.S. compared to last season, when much of the area was bombarded with heavy snow.

Risk management
Risk management is an advanced method for preseason snowfall event planning for any organization. This model calculates the statistical likelihood of the number of snow events and the amounts of snowfall that occur over a season. The definitive value of using a risk management model is that an organization will have the ability to plan the following winter without speculating the number of events that are typical and the probability of inches per event. This places organizations in a better position to manage risks and maximize opportunities presented by weather in a particular market.

Using the risk management model, this table gives the probabilities for one snowfall event.

This method is recommended for preseason planning:

• Budgeting – Planning seasonal vs. per-event or per-push contracts
• Procurement/Operations – Planning for equipment, materials, labor, snowfall insurance or dividends, etc.

Weather is the fundamental building block for constructing an actionable plan for the upcoming winter, and by speculating rather than using a statistical method you expose your organization to risk and missed opportunities every year.

The simple truth is that organizations that take a practical statistical approach to planning for any winter season will greatly improve its accuracy compared to those that just speculate. So please put away the “Magic 8 Ball” for 2011-12 and apply some business weather intelligence to give your organization the edge over your competition. 

Regardless of whether you are a small operation or a multimillion-dollar organization, there are weather tools available that can help mitigate the risk that comes from working in a weather-related industry. Below are some of the most common options with pros and cons listed for each to help you determine which might be right for your business.

Sean Hartnett, CSP, is a manager of client services for Planalytics. Contact him at 800-882-5881 or This e-mail address is being protected from spambots. You need JavaScript enabled to view it .

Back to "What If" Main Page