Four cities, one morning: the spread, not the median
On 6 October we read public Hostelworld prices in Lisbon, Porto, Barcelona and Oaxaca. Porto's cheapest priced bed sits 39.6% under its median; Oaxaca's sits 18.8%. The spread decides your move.
The dearest bed in Oaxaca costs less than the cheapest bed in Porto
At 08:12 UTC this morning, 6 October, our rate reader finished a pass over four cities and wrote down what the hostels in each were asking for a dorm bed. These are public Hostelworld prices, read off the pages by our own engine, one sampled night per city, one guest, in USD. Nothing here is modelled or estimated. It is what the listings said when we looked.
- Lisbon, night of 9 October: 30 listings seen, 7 carrying a price. Median USD 36.38. Cheapest 30.24. Dearest 55.69.
- Porto, night of 9 October: 30 listings seen, 11 carrying a price. Median USD 29.76. Cheapest 17.98. Dearest 65.63.
- Barcelona, night of 3 November: 33 listings seen, 8 carrying a price. Median USD 28.14. Cheapest 21.65. Dearest 41.31.
- Oaxaca, night of 9 October: 30 listings seen, 7 carrying a price. Median USD 12.13. Cheapest 9.85. Dearest 17.29.
Start with the line in the title. The dearest bed carrying a price in Oaxaca this morning was 17.29. The cheapest in Porto was 17.98. The top of one city's range sits below the bottom of another's, and Lisbon's median is three times Oaxaca's, near enough exactly. If you run in Oaxaca, as I do, none of that changes a rate you set this week. You do not compete with Lisbon. The number that moves your night is inside your own city, and it is not the median.
Porto and Lisbon: same morning, same night, different shape
Those two were read the same morning for the same night, so they go side by side without an asterisk. Lisbon has the higher median, 36.38 against 29.76, and on the medians alone it looks like the better room to be standing in. Then read the ends. Lisbon ran 30.24 to 55.69: the top is 1.8 times the bottom and the cheapest bed sits 16.9 per cent under the median. Porto ran 17.98 to 65.63: the top is 3.6 times the bottom and the cheapest sits 39.6 per cent under. Same country, same channel, same night, and one band is twice as wide as the other.
That is what changes your Tuesday. In Lisbon, dropping to the floor of the visible band costs about six dollars a bed. In Porto it costs about twelve, and somebody that night is holding a bed at 2.2 times the city median. A wide band means the market has not agreed what the bed is worth, so there is room to move either way. A narrow one means it has agreed, and the room you thought you had is mostly not there.
What a 25-bed property does with it
Put it in beds, because beds are what you sell. Take 25 beds and ask one question: priced at the floor of my city's band instead of its median, how many do I sell to finish the night on the same money? Ten beds at the median is the baseline.
- Lisbon: 10 beds at 36.38 is USD 363.80. At 30.24 you need 13. Three more beds.
- Barcelona: 10 at 28.14 is 281.40. At 21.65 you need 13. Three more beds.
- Oaxaca: 10 at 12.13 is 121.30. At 9.85 you need 13. Three more beds.
- Porto: 10 at 29.76 is 297.60. At 17.98 you need 17. Seven more beds.
On 25 beds, three more is 40 to 52 per cent occupancy. Seven more is 40 to 68. The first is an ordinary shoulder-season Tuesday; the second is a different business. In three of these four cities the floor costs you three beds a night. In Porto it costs seven, and whoever goes down there has to find twenty-eight points of occupancy that were not coming on their own.
Across a month at 60 per cent occupancy, 15 beds a night, the floor instead of the median is roughly 5,300 dollars less in Porto, 2,900 in Barcelona, 2,800 in Lisbon and 1,000 in Oaxaca. That is one morning's band held flat for thirty nights, which no month does. The ranking is the part that holds.
A narrow band is why a discount stops working
We have our own receipt here. The engine pricing our 46-bed hostel in Oaxaca was not handed a last-minute discount; it learned one, walking from 0 to minus 8 per cent in three-point steps, then stopping because its own read said deeper was not buying beds. Across 32 logged runs, days at minus 8 per cent averaged 33.7 per cent occupancy over 23 days; days at 0 per cent averaged 42.0 per cent over 5 days.
Five days against twenty-three, one property, one low season, no control for day of week. That is a hint, not a law, and I will not dress it up. But it points where the Oaxaca band points. In a city whose entire visible range this morning was 9.85 to 17.29, a discount is not a lever. It is 2.28 dollars. You will not out-cheap a market that is already cheap; you will reach the same occupancy holding less money.
The part I will not round off
Across the four cities we saw 123 listings and 33 carried a price for the sampled night. The other 90 did not. We record what came back; we do not know, listing by listing, whether that meant sold out, closed, not selling that bed, or simply no rate returned to a public page. Every median above is therefore a median of what was priced and visible this morning, on seven to eleven properties a city. A median of seven moves if one hostel changes its mind.
One more asterisk: Barcelona's sampled night was 3 November, the other three 9 October. A price a month out and a price three nights out are different measurements, so Barcelona belongs in the list but not in a head-to-head with Porto.
None of these prices is right or wrong, and I am not naming a property and what it charges. The useful question is narrower: where in your city's band are you standing, and is the band wide enough for the move you were about to make?
Doing this for your own city
- Fix the spec and keep it: one night, one guest, the bed you actually sell, the channel that sends you volume. A reading taken differently each week measures your method, not your market.
- Write down how many listings carried a price, not just the median. Seven of thirty is a thinner fact than eleven of thirty.
- Record cheapest, median and dearest, then place yourself between them. The median says where the market sits; the distance between the ends says whether price is a lever you have.
Nightfill does this work as a service rather than another tool you have to run; the tiers are on the pricing page.
If you want this morning's reading for your own city rather than these four, paste your Hostelworld listing at nightfill.co/#read and the same reading for your property arrives by email in ten minutes.