FORECASTING
by Victoria Edwards


Forecasting is the cornerstone of any good revenue management program. Yet for the most part it is not a core competency in our industry. The benefits of a forecast are clear to all - if you are able to accurately predict the volume and type of business coming into your marketplace, you can optimize both revenue and profit for your business. 

If I were a general manager in today’s highly competitive and volatile markets, I would ensure that my revenue management team had the necessary training and tools to accurately forecast business. After all, forecasts are the starting point to both strategic pricing and strategic distribution decisions. 

Not all forecasts are created equal. Our industry utilizes four distinct types of forecast methodologies: occupancy, revenue, demand and fair share. Each one of these methodologies has a very different purpose. Occupancy forecasting is an operations-oriented constrained method used to assess short-term labour and purchasing needs. 
By contrast, revenue forecasting is much more financial in orientation. It is often done on a rolling 30/60/90-day basis and is used for such things as cash flow projections and decisions about larger non-capital purchases. It also is a constrained method of forecasting.

For the purpose of revenue management, the demand forecast and fair share forecast are the most critical in optimizing revenue and flow through. Fair share forecasting is a relatively new approach that very few hotels embrace. This is likely owing to the fact that fair share forecasting is much more complex than the other three methods, and although valuable, yields less accurate results.

So let us turn our attention to the most critical of the four methods - demand forecasting. This is the most important of the methods because if you do not have an accurate demand forecast, both your occupancy and revenue forecasts will also be inaccurate.

The key to demand forecasting is that it is based on unconstrained demand. That is, it is based on the total amount of measurable demand coming into your marketplace. Two of the ideal conditions for practicing optimal revenue management are fixed capacity and a perishable inventory. We have fixed capacity since we are not able to manufacture more rooms based on short-term demand. Additionally, in the case of hotel rooms, the sales opportunity of a given day is lost every 24 hours. It perishes on a daily basis if it remains unsold. We cannot resell an unoccupied January 1st room on January 2nd. 

Unconstrained demand goes beyond fixed capacity, and essentially says, “If you had an unlimited amount of rooms in your inventory, how many could you sell?” In other words, it goes beyond 100% occupancy. In very high demand periods a hotel may have the ability to sell 120% of its inventory. In super compression, such as the 2010 Olympics in Vancouver the demand for your hotel will significantly outpace supply both within your competitive set as well as the marketplace in general. 

Clearly, knowing whether the demand for your hotel is 70%, 105%, or 130% will impact not only the way you sell your product but also the channels through which you choose to distribute your product. You will make pricing decisions based on this information. After these pricing decisions are made you will want to leverage overall contribution by being selective about your distribution choices. All channels have some costs associated with them, and in the case of exceptionally high demand, you can afford to be more selective about the amount and type of product that is offered in each channel.

In order to build your hotel demand forecast, you will need to look at a number of variables. These variables need to be tracked on a granular basis. In other words, you will need to look at your business from a micro-market segment basis. The more granular your data becomes, the better the information to make pricing decisions. The variables that should be addressed are:
• Reservations on the books
• Current daily pace by market segment
• Pace over previous year by market segment
• Turndowns - inventory and price-based
regrets/denials as well as inquiries
• Demand influencers - both negative and
positive as they pertain to your market place
as well as specifically to your competitive set
This information should be gathered on a daily basis and will formulate the basis of your demand forecast. An important note should be made on the weight we give to historical data. Certain recurring events will always have an impact on your forecasts. Events such as the Calgary Stampede in Alberta or Carribana in Toronto will continue to provide predictive information to a demand forecast. However, as booking windows continue to shrink, we need to become less reliant on overall historical booking patterns when shaping our decisions. History is still a part of your forecast, but needs to be weighted somewhat differently. The big revenue management software companies used to data-mine 12 months worth of booking patterns in order to establish rate hurdles during software installations. They have found that this amount of data is far less predictive than 90-120 days.

Once this information has been gathered daily by micro-segment, you can then start to forecast unconstrained daily demand. This in turn will give you the information you need to forecast total occupied rooms, and then create a revenue forecast based on your pricing decisions - pricing decisions based on the balance you want to achieve in your business mix to fully leverage both rate and occupancy while protecting your most valued customers.

Just like any other aspect of the revenue management discipline, a proper forecast practice follows a business process. The ideal process includes:
1. Assessing Forecast Accuracy - Look at accuracy on a daily rather than monthly basis, and analyze by market segment. Studies indicate that a 10% improvement in forecast accuracy results in a 1%-2% increase in top line revenues.
2. Big Picture - Make sure the team is discussing issues that may impact the forecast. Discuss the specifics of short, medium and long-term demand influencers. These influencers may be positive or negative, and may be locally, regionally, or nationally relevant.
3. Trends Analysis - Make sure your team is getting the most out of both competitive intelligence and what the forecast accuracy report is telling them. Apply lessons learned from over or under-forecasting of specific segments or day of the week patterns.
4. Output Documents - Ensure that you are forecasting demand before you attempt an occupancy or revenue forecast.

The forecast process needs to be a living thing that involves the entire revenue management team. Each member of the team is a subject matter expert. The reservation manager can give a very clear picture with regard to what is and is not selling. Your front office manager can offer valuable information in terms of arrival/departure patterns, no-show factors, and other valuable pieces of the puzzle. Clearly, each of your sales managers needs to fully understand their segments, and be accountable for wash in their segments.

Finally, the director of sales and marketing needs to continue his leadership role by working closely with the director of catering/conference services. As the discipline of revenue management continues to evolve, decisions need to be based on maximizing all aspects of the business asset. This means integrating rooms with food and beverage revenue management so that RevPASH (revenue per available seat hour) and RevPSF (revenue per square foot of catering space) are also optimized.

Catering space demand forecasting is an area that very few hotels have focused on. Demand forecasting in this area tends to be much more complex as you are dealing with more variables. Catering space has somewhat variable capacity, given the multiple options available for room configurations (classroom, rounds, theatre style, etc.). Additionally, while rooms are sold on a day basis, function space needs to consider day segments such as breakfast, coffee break, lunch, dinner, and reception. Multiple day segments and variable capacity make defining occupancy somewhat more challenging. However, demand forecasting in the domain of catering is not impossible, and will likely become the next frontier of the highly integrated revenue management team.

The key to demand forecasting is to treat it as a business process. Ensure that your team has the training and tools that it needs to do the job. The revenue management team needs to make decisions, both pricing and distribution decisions, in an integrated and holistic manner. In doing so, the team will be better able to approach the business from both a tactical and strategic perspective.

Victoria Edwards is the VP Strategic Development & Co-founder of Buckhiester USA Inc. She is the creator of the RevRoadMap business process and diagnostic process as well as the co-creator of REVolution, a complete resource center of tools, procedures, workshops, and coaching in revenue management. www.buckhiester.com