Every seasonal property faces the same annual decision, and most make it the same way. Demand falls, rooms sit empty, fixed costs continue, and the property discounts. The logic feels sound: some revenue is better than none, and a discounted room still contributes toward overhead.
The logic holds only if the demand exists and is price-responsive. In a meaningful proportion of seasonal troughs, neither condition is met, and the discount destroys value in both the trough and the peak.
Testing whether the demand is there at all
Before a resort discounts into a trough, it should establish whether the trough is a price problem or a demand problem. The test is straightforward: plot historical occupancy against realised ADR for the trough window across several years, and look for a relationship.
For an 84-key Goa resort we examined, five years of monsoon-season data showed ADR falling 51% from the annual mean while occupancy remained below 20% throughout. There was no discernible elasticity. The property had run the experiment repeatedly and unintentionally, and the answer each time was that the guest who wanted a beach holiday in heavy rainfall did not exist at ₹6,500 and did not exist at ₹3,200 either.
| Year | Monsoon ADR | Discount vs annual mean | Monsoon occupancy |
|---|---|---|---|
| Y-5 | ₹4,100 | -37% | 21.4% |
| Y-4 | ₹3,760 | -42% | 19.8% |
| Y-3 | ₹3,340 | -49% | 18.2% |
| Y-2 | ₹3,210 | -51% | 19.6% |
| Y-1 | ₹3,180 | -51% | 19.0% |
Monsoon-window occupancy against realised ADR, 84-key Goa resort, five-year history
Four consecutive years of deepening discount produced no occupancy response whatsoever. The property was giving away 51% of its rate in exchange for nothing measurable.
The cost that lands in the peak season
Trough discounting is rarely contained to the trough. Guests anchor on the lowest rate they have seen a property advertise, and the marketplaces that surfaced the discounted rate retain it in cached results, price-history features and rate-alert notifications for months afterward.
The practical consequence is measurable resistance at peak rates. Front-office teams at the property in question reported recurring rate objections in the December peak that referenced monsoon pricing directly. A guest who has seen ₹3,200 does not readily accept ₹9,000, regardless of how well the seasonality is explained.
There is a second-order effect that is harder to see and more damaging. Marketplace ranking algorithms classify inventory that is persistently discounted as discount inventory, and surface it preferentially in price-ascending sorts. The property in question appeared in price-sorted results on 87% of sampled searches but in relevance-default sorts on only 22% — a structural suppression of full-rate demand that persisted into peak season.
Repositioning rather than repricing
The alternative is to change what is being sold rather than what it costs. The resort in question stopped selling a discounted beach holiday during the monsoon and started selling two different propositions to two different guests.
The first was wellness: five- and seven-night packages bundling spa treatments, in-resort dining credit and late checkout, sold at rate integrity to guests for whom the rain was irrelevant because the proposition was indoor and restorative. The second was work-from-resort: extended stays targeting metro professionals, led on connectivity, workspace and length-of-stay value rather than on beach access.
Neither proposition required capital expenditure. Both required the property to stop describing itself in monsoon months using peak-season imagery and peak-season language, and to stop competing for a guest who was not travelling.
| Metric | Prior monsoon | Post-repositioning | Movement |
|---|---|---|---|
| Occupancy | 19.0% | 51.0% | +32.0 pts |
| ADR | ₹3,180 | ₹5,940 | +86.8% |
| RevPAR | ₹604 | ₹3,029 | +401% |
| Average length of stay | 2.3 nights | 4.1 nights | +78.3% |
| Share of stays 5+ nights | 4% | 38% | +34 pts |
Monsoon-window performance, prior year against post-repositioning year
Why length of stay matters more than rate
The most under-appreciated figure in that table is not RevPAR. It is average length of stay moving from 2.3 to 4.1 nights.
Length of stay changes resort operating economics in ways rate does not. Housekeeping cost concentrates on arrival and departure days; a four-night stay costs materially less to service per room night than two two-night stays. Front-office labour, arrival transport and check-in processing all follow the same pattern. In-resort F&B capture rises with stay length because guests who are settled explore the property's own outlets rather than eating out.
A property that lengthens stays improves its margin structurally, at the same rate and the same occupancy. Very few trough strategies deliver that; almost none of them are discounting strategies.
Before you discount into a trough
- Plot several years of trough ADR against trough occupancy. If there is no relationship, the demand is not price-elastic and the discount buys nothing.
- Ask what proposition would make the trough window attractive to a different guest, rather than what price would make it attractive to the same one.
- Model the peak-season rate resistance your trough rate creates before committing to it.
- Target length of stay explicitly. It improves margin at constant rate and occupancy, which no discount can do.
This analysis draws on engagements led by the author. Findings are anonymised at client request; underlying figures are taken from client property management, channel manager and advertising platform records.
