BUSN Case Study 2: Intercontinental Hotel Room Forecasting

A hotel room is a perishable good. If it is vacant for one night, the revenue is lost forever.
Monica Alvarez was commenting on the issue of capacity utilization in the hotel business. On
the other hand, the customer is king with us. We go to great pains to avoid telling a customer
with a reservation at the front desk that we dont have a room for him in the hotel.
As Manager of Revenue Managment at Intercontinentals hotels, Monica faced this tradeoff constantly. To complicate the matter, customers often booked reservations and then failed to
show, or cancelled reservations just before their expected arrival. In addition, some guests stayed
over in the hotel extra days beyond their original reservation and others checked out early. A key
aspect of dealing with the capacity-management problem was having a good forecast of how
many rooms would be needed on any future date. It was Monicas responsibility to prepare a
forecast on Thursday afternoon of the number of rooms that would be occupied each day of the
next week (Saturday through Friday). This forecast was used by almost every department within
the hotel for a variety of purposes; now she needed the forecast for a decision in her own
department.
Times Square Hotel
The Times Square Hotel was a large downtown business hotel with 1220 rooms and
abundant meeting space for groups and conventions. It had been built and was operated by
Intercontinental Hotels, a company that operated more than 4600 hotels and resorts in more than
90 countries worldwide and was expanding rapidly into other lodging segments. Management of
The Times Square reported regularly to Corporate headquarters on occupancy and revenue
performance.
Hotel managers were rewarded for their ability to meet targets for occupancy and
revenue. Monica could not remember a time when the targets went down, but she had seen them
go up in the two years since she took the job as manager. The hotel managers were continuously
comparing forecasts of performance against these targets. In addition to overseeing the
reservations office with ten reservationists, Monica prepared the week-ahead forecast and
presented it on Thursday afternoon to other department managers in the hotel. The forecast was
used to schedule, for example, daily work assignments for housekeeping personnel, the clerks at
the front desk, restaurant personnel, and others. It also played a role in purchasing and revenue
and cost planning.
Overbooking
At the moment, however, Monica needed her forecast to know how to treat an
opportunity that was developing for next Saturday. It was Thursday, July 17, and Monicas
forecasts were due by midafternoon for Saturday, July 19 through Friday, July 25. Although 1195
rooms were reserved already for Saturday, Monica had just received a request from a tour
company for as many as 250 more rooms for that night. The tour company would take any
number of rooms less than 250 that Monica would provide, but no more than 250. Normally
Monica would be ecstatic about such a request: selling out the house for a business hotel on a
Saturday would be a real coup. The request, in its entirety, put reservations above the capacity of
the hotel, however. True, a reservation on the books Thursday was not the same as a head in the
bed on Saturday, especially when weekend nights produced a lot of no-show reservations.

Chances are good we still wouldnt have a full house on Saturday, Monica thought out loud.
But if everybody came and someone was denied a room due to overbooking, I would certainly
hear about it!
Monica considered the trade-off between a vacant room and denying a customer a room.
The contribution margin from a room was about $180, since the low variable costs arose
primarily from cleaning the room and check-in/check-out. On the other side, if a guest with a
reservation was denied a room at The Times Square, the front desk would find a comparable
room somewhere in the city, transport the guest there, and provide some gratuity, such as a fruit
basket, in consideration for the inconvenience. If the customer were an IHG Gold cardholder (a
frequent guest staying more than 30 nights a year in the hotel), he or she would receive $100
cash plus the next stay at Intercontinental free. Monica wasnt sure how to put a cost figure on a
denied room; in her judgment, it should be valued, goodwill and all, at about twice the
contribution figure.
Forecasting
Monica focused on getting a good forecast for Saturday, July 19, and making a decision
on whether to accept the additional reservations for that day. She had historical data on demand
for rooms in the hotel; Exhibit 1 shows demand for the first 3 weeks for dates starting with
Saturday, April 19. (Ten additional weeks [weeks 4-13] are contained in Intercontinental.xlsx and
thus Saturday, July 19, was the beginning of week 14 in this database.)
Exhibit 1
Historical Demand and Bookings Data

Demand figures (column C) included the number of turned-down requests for a reservation on
a night when the hotel had stopped taking reservations because of capacity plus the number of
rooms actually occupied that night. Also included in Exhibit 1 is the number of rooms booked
(column D) as of the Thursday morning of the week prior to each date. (Note that this Thursday
precedes a date by a number of days that depends on the date's day of week. It is two days ahead
of a Saturday date, seven days ahead of a Thursday, eight days ahead of a Friday. Also note that
on a Thursday morning, actual demand is known for Wednesday night, but not for Thursday
night.)
Monica had calculated pickup ratios for each date where actual demand was known in
Exhibit 1 (column E). Between a Thursday one week ahead and any date, new reservations were
added, reservations were canceled, some reservations were extended to more nights, some were
shortened, and some resulted in no-shows. The net effect was a final demand that might be larger
than Thursday bookings (a pickup ratio greater than 1.0) or smaller than Thursday bookings (a
pickup ratio less than 1.0). Monica looked at her forecasting task as one of predicting the pickup
ratio. With a good forecast of pickup ratio, she could simply multiply by Thursday bookings to
obtain a forecast of demand.
From her earliest experience in a hotel, Monica was aware that the day of the week
(DOW) made a lot of difference in demand for rooms; her recent experience in reservations
suggested that it was key in forecasting pickup ratios. Downtown business hotels like hers tended
to be busiest in the middle of the workweek (Tuesday, Wednesday, Thursday) and light on the
weekends. Using the data in her spreadsheet, she had calculated a DOW index for the pickup
ratio during each day of the week, which is shown in column F of Exhibit 1. Thus, for example,
the average pickup ratio for Saturday is about 86.5% of the average pickup ratio for all days of
the week. Her plan was to adjust the data for this DOW effect by dividing each pickup ratio by
this factor. This adjustment would take out the DOW effect, and put the pickup ratios on the
same footing. Then she could use the stream of adjusted pickup ratios to forecast Saturday's
adjusted pickup ratio. To do this, she needed to think about how to level out the peaks and
valleys of demand, which she knew from experience couldn't be forecasted. Once she had this
forecast of adjusted pickup ratio, then she could multiply it by the Saturday DOW index to get
back to an unadjusted pickup ratio. "Let's get on with it," she said to herself. "I need to get an
answer back on that request for 250 reservations."
Questions
1. Verify the Day-of-Week indices in Column F of Exhibit 1. If you dont agree with Monica,
give your new indices.
2. What forecasting procedure (try at least 3) would you recommend for the adjusted pickup
ratio? Report the same error measure (e.g., MAD, MAPE, or MSE) for each method tried.
3. What is your forecast for Saturday, July 19? What will you do about the current request for up
to 250 rooms for Saturday?