The most expensive month you will have this year is one in which everything sold. On the one-bed worked through below, running the calendar full instead of pricing it properly costs about R2 850 a month, near enough R34 000 a year, and nothing on your dashboard reports it as a loss. That is the trouble with Airbnb occupancy in Cape Town: it is the number every host quotes at the braai and the number least connected to how much money they made. If your self-catering unit is occupied thirty nights a month, you have not discovered a magic listing. You have discovered the price at which nobody says no, and the price at which nobody says no is, by definition, below the price at which some people would still say yes. Full is comfortable. Full is not the same as paid.
The maths of fewer bookings
Consider a one-bed charging R1 000 a night and booking all thirty nights of a month. Gross: R30 000. Now suppose the same flat asks R1 400 and a quarter of the guests walk away. Occupancy drops to 75 percent, which is 22 or 23 nights, and gross rises to R31 500. You have earned more money from fewer guests, and the story improves once you count what did not happen. At an average stay of two and a half nights, that is three fewer turnovers at R450 each, R1 350 you did not spend on cleaning. Three fewer sets of linen through the laundry, three fewer check-in messages, fewer chances of the guest who leaves a two-star review because the Wi-Fi hiccupped. The unit sits empty a week a month, which is also a week in which nothing breaks.
Hoteliers have a word for the score that settles this: RevPAR, revenue per available night. It is the only metric that cannot be gamed by being cheap, and almost no host tracks it. Your full calendar has a RevPAR of R1 000. The quarter-empty one next door, asking R1 400, has a RevPAR of R1 050 and a much quieter washing machine.
Why full feels like winning
The instinct is understandable. Empty nights are visible, each one feels like a small failure, and thirty bookings in a row feel like proof of competence. But the calendar only shows one side of the trade. It shows the guest who booked at R1 000. It does not show the guest who would have booked at R1 400 and never had the chance, because the night was already gone to someone who paid less. Underpricing has no symptom. That is why the hosts who do it are so confident: their evidence is a full calendar, and a full calendar is exactly what underpricing produces.
What Airbnb occupancy in Cape Town is measured against
Benchmarking your way out does not work either: the published Cape Town numbers disagree with each other. As we write, in mid-September 2026, one published market report puts occupancy at 43 percent and another at 71, while our own market snapshot from three weeks earlier showed 67 percent city-wide and 71 percent for the City Bowl on its own. They are all defensible: they divide by different calendars, as we unpacked in Occupancy: the slowest number in hosting. The practical consequence: a host at 95 percent looks like a genius against every one of those benchmarks, and not one of them says a word about rate. You can beat the market on occupancy every month of the year by being the cheapest unit on your street. You will also earn less than your street.
Where the balance point is
The question is not "should I raise my price" but "how far, and on which nights." Every listing has a point where the next R100 on the rate loses more nights than it gains in revenue. That point moves with the season, the day of the week, the events in your suburb and how many comparable units are still open, which is why it cannot be found once and left. In December the balance point for a Sea Point one-bed sits far above where fear puts it, because supply runs out. In July it sits close to the floor, because the guests do. On a Kirstenbosch Sunday in Newlands it jumps for one night and falls back on Monday.
That last case is the one hand-pricing always loses. The balance point does not move on the day of an event, it moves on the day the event is announced, because guests planning around a race number or a concert ticket book immediately. That is the job we built for: RateVeld keeps a hand-curated Cape Town events calendar wired into the pricing curve, and affected dates reprice on announcement day, six months out, in rand, with the reasoning shown against each night. It carries the national school calendar and the ISASA private-school calendar too, because Gauteng's private schools break for spring two weeks before the public ones and return a week later in autumn, and a guesthouse priced to only one of those calendars gives away one end of its own holiday. What pricing to events instead of through them is worth depends on your suburb, your season and how you price today, so we will not promise you a figure; the shape is winter nights that hold their floor and event weeks that more than double.
Finding the balance point by hand means reading comparable listings and hotel rates for every open date, every week, and noticing the morning an event is confirmed. Most hosts price hard in March, again in November, and call the ten months between a strategy. And if your objection is that your suburb is different, that argues for this rather than against it: the balance point in Muizenberg is not the balance point in Green Point, which is why a model calibrated on South African market data beats one that learnt its seasonality in Florida.
The question "how to price my Airbnb" is usually asked as though the answer is a number. It is not. It is 365 numbers, revised whenever the evidence changes. Dynamic pricing for Airbnb is, at bottom, a way of running that experiment on every night at once, using what comparable listings and nearby hotels are asking as the evidence, so the balance point is estimated per date rather than discovered per year.
Comfort has a price
Some hosts will read all this and decide they prefer the full calendar anyway: the certainty, the steady rhythm, the absence of dark nights on the app. That is a legitimate choice, as long as it is made knowingly. Pricing for comfort costs, in the example above, the R2 850 a month we opened with: about R34 000 a year on one small flat, for the pleasure of never seeing a gap. If it is not worth that to you, stop reading occupancy as a report card and start reading it as a signal: a calendar that is always full is the market telling you it would pay more. How much more, and on which nights, is the part that takes evidence.
What members get
RateVeld finds the balance point per night rather than per year. It reads your suburb's comparable listings and nearby hotel rates every day, prices every night for the next six months in rand to maximise revenue rather than occupancy, and shows the reasoning behind every rate. It never recommends below the floor you set, and it never chases a full calendar for its own sake. A short term rental pricing tool should make you money, not make you busy.
Plans are priced in rand, per property. See the plans and prices and pick the one that fits your calendar. If the full calendar is a December one, read the December rate Sea Point hosts are scared to charge next.