Demand forecasting: where to start when you have no tool
A well-built spreadsheet and two hours a week are enough to lay the foundations. The tool will come later, if it is justified.
Many hotels put off forecasting until they have “the right software”. It is the reverse order that works: start by hand, understand your own demand, and only then choose a tool, knowing what you expect from it.
The three data series to rebuild
- The pickup curve: how many rooms had been sold 60, 30, 14 and 7 days before arrival for each date last year.
- The demand calendar: events, conferences, school holidays in source markets, public holidays.
- The channel and segment mix, month by month.
All three series can be extracted from a PMS, even an old one. Formatting them takes a day. The work will serve you for years.
The weekly routine
Once a week, at a fixed time, compare your position against the reference curve for the next four to six weeks. Note the gaps, decide, and write the decision down. This last point is the most important: without a written record, nobody can assess whether past decisions were the right ones.
“Revenue management that leaves no written record never improves: you cannot correct decisions you no longer remember.”
After a quarter, you will know which dates are genuinely in high demand, which segments book late, and where demand is elastic. That is when a forecasting tool becomes an accelerator — rather than an expensive gadget.
