The first method called Same-day-last-year uses previous year data to forecast this year bookings. The model assumes that the demand is similar when we take the same day in previous year, remembering about order of the week. For example, demand on Monday, first week of July 2013 should be the same as Monday, first week of July 2012 (Phumchusri and Mongkolkul, 2012). This method is easy to compute while on the other hand is very sensitive to daily special events in the past.
The advantage of this method is fact that we can actually forecast the entire year ahead (the forecast in the spreadsheet is conducted for 2013)
This method might be expanded and instead of using the last year value, we can calculate the average of bookings in certain day in couple of years before.
The application of this method on our data set is available here: Same-day-last-year
1st January 2013 is Tuesday, therefore we need to take bookings from 3rd January 2012 (as it also is Tuesday)
When demand is characterized by significant daily variation, this method will give poor results. In the spreadsheet attached above, the MAPE = 8% and 7%. On the real dataset, the MAPE=41% and also the graph shows the inefficiency of this method.
Phumchusri, D., Mongkolkul, J. (2012) Hotel Room Demand via Observed Reservation Information. Proceedings of the Asia Pacific Industrial Engineering & Management Systems Conference 2012, pp. 1978-1985