The end of the 2011-12 ski season was met with numerous reports of how bad the year was. Most of these statements were made in comparison to the superb 2010-11 ski season. There is no argument that last season was a difficult one in most areas of the country, but the year-to-year comparison gives a distorted picture.

Upon reflection, I began to question if this year-to-year comparison is the right metric. Growth is a given metric most all businesses strive for, but in a weather-dependent business, what is growth?

In simplistic terms, growth means total revenues grew above the previous year’s revenue. But perhaps a more meaningful measure would be to tally skier days and revenue against a normal year’s performance.


BUDGETING FOR NORMAL
I can hear the shouts: “What’s normal?” My response: Most ski resorts report average snowfall—that is one measure of normal. Now, what are your average skier visits associated with that snowfall? And the average revenue? That’s your normal year.

To my knowledge, this is not a typical baseline ski areas use to measure themselves, but maybe it should be.

Let’s first understand how we might calculate the baseline. It could be very simple, or more complex and take some significant effort to calculate (see boxes for arguments in favor of each approach).

For the simple approach, look at the last 10 years of snowfall. Identify if any anomalies of a low-snow or great snow year exist in that time frame. If they do, review your operation for that timeframe and determine if you made any significant changes, such as a major expansion of lift capacity or terrain, and identify the skier day growth associated with that expansion, apart from changes in snowfall. If growth has been consistent since the expansion, shorten the time frame to include only those years since the expansion. (If this period includes fluctuations in snowfall and visitation, it will more accurately reflect your true business.) Then average the skier visits for the 10 years (or the adjusted timeframe) and this becomes your baseline for visits going forward. Simple!

A more complex method is to build a model based on history, by day, putting weather (snowfall and conditions), visits and revenue into the model (below). Mad River Glen, Vt., has been doing this for 17 years, using a ranking system for the daily conditions, with 1 being poor and 5 being super. This model has the ability to forecast daily expected visits and revenue, adjusted for time of year, for any given 1 to 5 ranking. The model has been a useful tool over the last decade for managing revenues and expenses.


MAD RIVER GLEN'S MODEL
When the Cooperative purchased Mad River Glen in 1995, the incoming chairman of the board asked the assistant GM (now president), what is the relationship of revenue to snow? The answer was as expected for an area with next to no snowmaking, and so began the development of a model which measures that relationship between snow and revenue. This model has become a very valuable tool.

From day one, the Coop began coding snow conditions each day on a 1 to 5 scale (1 being rain with one lift operating, and 5 being two feet of new snow with all lifts spinning) and recording the paid ticket revenue associated with that day. This coding has been done by the same person from the beginning, providing consistency in the ranking of each day. The 17 years of history reflect many variations of weather and events beyond management’s control, from the ice storm of 1998 to the loss of the bearing on the single chair’s bull wheel over a Christmas week.

The model has been refined over time. It omits days when operations are suspended due to conditions, and the conditions rating is done to the tenths to provide better clarity on the quality of surface conditions when there is no fresh snow. The coding gets somewhat adjusted for big events, such as the annual NATO Telefest, which blunt the impact of bad weather.

The real strength of this model is that the input is straightforward and simple. (The model itself is less simple; it was developed by a young college grad who was waiting his opportunity with a large financial management house.) It is both a strength and weakness that one person has done the conditions rating, and that it is somewhat subjective. Mad River plans to address this by making the conditions coding more objective, so that it can be done by others without distorting the data.

The use of the tool allows management to know, based on the average rating, what revenue to realistically expect on a daily, weekly, monthly or annual basis. The model accounts for variations by month, week and day, as well as the variable of pricing, one of the few aspects management can control.

It only fails when the weather acts in ways beyond anything that has happened during the period that data have been recorded. This happened during March and April of 2001, when snowfall (and revenue) was off the charts. That skewed the model for a couple of future March and April periods, but time has minimized that impact.

Here’s an example of the daily matrix over a two-week period, including a holiday week:

Day of Week Snow Condition Code/Ticket Revenue Average Revenue

1 2 3 4 5

Sunday 1,282 7,212 15,059 26,492 38,428 16,701.90

Monday 539 2,511 4,675 7,315 11,679 4,273.49

Tuesday 653 1,975 3,999 7,074 11,679 3,836.92

Wednesday 665 1,945 3,798 7,859 11,679 3,894.23

Thursday 521 1,757 3,388 6,072 11,771 3,352.18

Friday 1,035 4,010 9,486 14,759 22,043 9,778.74

Saturday 2,541 8,449 20,451 31,915 51,577 22,986.60

No model is perfect, but this example demonstrates how a small ski area developed a meaningful relationship to snow and revenue, and uses it to manage its business. It is not making guesses or forecasting arbitrary levels of visitation. Mad River Glen knows its capacity and its volume in relation to snowfall and surface conditions. How well do you know yours?

As an example of the simpler technique, consider the data (below) from the state of Vermont for 2011-12 prepared by the Vermont Ski Areas Association. The data have no relation to any specific resort, but provide a view toward the industry nationally, and serve as a good example for resorts large and small.

Here are the Vermont 2011-12 comparisons with the 10-year normal and prior year:

Normal Year Prior Year

Snowfall -34.7 % -43.9 %

Skier Visits - 6.0 % - 9.3 %

Clearly, there is a difference between one metric versus the other. The purpose is not to make one feel better when looking at a down year, but to use the “normal year” metric to help us better manage our operations over the course of a year.

Here’s how we might use this metric to forecast visits for Vermont for an upcoming winter:

Baseline skier visits 4,150,000

Projected growth in visits from:

5th and 6th grade programs 16,000

Learn to Ski & Ride Program 25,000

New Children’s programs 10,800

Down Country promotion 120,000

Total forecasted visits 4,321,800 (4.2 percent above normal)

The important aspect is not the forecast, but identifying measurable growth numbers. If it snows more or less than the average and that snow happens on prime dates, expect above-baseline skier visits not associated with any identified growth programs. If it snows less than average and snow conditions are not as the customer expects, then adjust operations to fewer skier visits.

This normal-year approach isn’t perfect, but unless you have the history and the ability to create a model such as Mad River’s, the simple averaging technique can serve you well in understanding what your baseline visits should be year-over-year in relationship to your snowfall. Accuracy will increase if you identify those programs or aspects of your operation that you can change to impact that baseline.

Just looking at year-over-year and picking an arbitrary number is not very smart. And a budget is only a guideline; as the snow falls you have to make adjustments. Accountants don’t often favor a rolling forecast/budget, but it will help you manage the bottom line.

I am sure that folks began to use this type of tool during the snow drought last year. Many were forced to ask, “what if it does X, Y or Z in terms of snowfall, what will we be looking at?” I would argue this should be a regular part of your management process. At least monthly, adjust your visit forecast to reflect what has happened to date, business trends, and the long term forecast in relation to your average snowfall for the balance of the season. Where does that take you? What adjustments should you be making? What marketing programs can be put in place to negate a problem or take advantage of a positive development? You know the drill. But making those decisions with better-formatted data can give you confidence that you are making the best decision possible.

Compared to a normal year, Vermont’s performance in 2011-12 was bad, yes (see chart above), but not as bad as everyone was screaming about. In fact, it was better than 2006-07, with 39 percent less snowfall than that season. Good job, Vermont! Declining just 6 percent from normal is something to be proud of in what everyone said was the worst year ever. And frankly, if we can’t take a 6 percent hit in a bad year, then maybe we have a bigger problem than snowfall.

Bob Ackland is a principal of Steep Management LTD. He was president of Sugarbush Resort, 2001-08, and CEO of Mad River Glen, 1999-2001.