Somewhere in your building there is a flat much like yours earning R150 000 a year more than yours, and its owner is not working harder than you are. Two one-beds in the same Sea Point block: same floor plan, same sea glimpse off the balcony, same body corporate, same cleaner some months. One clears about R380 000 in a year and the other about R530 000. Anyone who has run two units in one building has watched that gap open and gone looking for the visible difference: the photos, the sofa, the review count. Those matter at the margin. The decisive variable in Airbnb pricing in South Africa is not the decor. It is that one host prices every night on its own evidence, and the other set a rate in 2023 and hopes.
Same product, two rate cards
Call them Host A and Host B. Host A charges R1 400 a night, R1 700 at weekends, and raises it to R2 200 for "December" on 1 December. That is a three-number rate card, and to the person running it, it feels like dynamic pricing. Host B runs a curve: every night has its own rate, set on its own evidence, so a July Tuesday, an October weekend, the Christmas core and New Year's Eve are four different products at four very different prices, with sharp one-off spikes on the nights the city is busy. Host B's realised average is higher, but not dramatically so. The difference is in which nights were sold at which number.
Look at where that R153 000 comes from, because it is not one place. Roughly a third is December: Host A's R2 200 sold out in October to guests who would have paid far more in the Christmas core, and Host A never saw the money leave, because the calendar was full. Roughly a third is winter: Host A's R1 400 sat too proud on midweek nights in June and July while Host B's sold, and the empty nights read to Host A as a slow season rather than a wrong price. The last third is the twenty-odd nights a year that were worth more than their season, because something was happening in the city, and Host B moved on the announcement while Host A moved on nothing.
What Airbnb pricing in South Africa actually rewards
Lazy is an unkind word for what Host A is doing, so let us be fair about the mechanism. Host A is running on a feedback loop that closes once a year. December looks fine in December and only reveals itself in January, when the money is counted, and the correction can only be tested twelve months later. Winter looks like a soft market rather than a proud price, and there is no A/B test running in the flat next door to say otherwise, except that there is, and it is Host B. We set out the loop in Occupancy: the slowest number in hosting: four lessons a year against a market that offers 365.
Host A is also leaning on the weekend premium as the whole of their dynamic pricing, which is the most common pattern we see among hosts running one to five listings. Weekends do cost more than weekdays. But a Saturday in July and a Saturday in December are not the same Saturday, and a Saturday in March with the Cycle Tour finishing down the road is not either. A single weekend uplift captures the smallest of the three layers and misses the two that carry the money.
The third layer is the one that punishes hand-pricing hardest, because it moves on days nobody has diarised. A race date, a festival line-up or a conference lands in the news in September and sells out the Saturday in question long before the week it falls in. By the time Host A notices, the rooms are booked at a July number.
Where the reviews come in
There is a second-order effect that widens the gap over time and makes it look like the decor after all. Host B, pricing fairly on quiet nights, sells more of them to guests who feel they got value, and collects reviews at a steadier rate through the winter, which feeds ranking into the summer. Host A, proud in winter, collects fewer, and by year three the flat that looked identical in year one has a visibly shorter review history. Host B can then charge a little more on the same night and still convert first, and Host A concludes that Host B's flat must be nicer. It is not nicer. It has been priced with more attention for longer, and reviews are how the market remembers.
Airbnb occupancy in Cape Town, and what an average hides
This is not a Sea Point curiosity. The City Bowl on its own appears in our latest world-markets snapshot as a market ranked 49th on the planet by listings, with 4,789 of them, 71 percent occupancy and an average booked night of R1 836. One Cape Town district, out-scaling entire European cities. Inside a district with those averages, a flat that is physically interchangeable with the one upstairs can be worth R1 100 on a wet Wednesday in June and R3 400 on a race weekend in March, and the district average launders every bit of that information out. Two hosts orbiting the same R1 836 can finish R150 000 apart in a year without either of them ever being the cheapest or the dearest flat on the street. Airbnb occupancy in Cape Town is not the score. Occupancy multiplied by the rate you got for each of those nights is the score.
Why Host B is hard to copy
Being Host B is not a talent. It is a workload. It means knowing, for every night in the next six months, what the comparable listings on your street are asking for that specific date and how many of them are still open; what the hotels down the road have moved to; which school calendar that night sits inside, the national one or the independent one; and whether an event announced this morning has just turned an ordinary Saturday into the most valuable night of your autumn. Then it means doing it again tomorrow, because all of that moved overnight. Hosts who try it by hand do it for a week, keenly, and then December arrives and the linen wins.
And the part of the gap that hurts most is the part you cannot see in your own records. The December third shows up, eventually, as a calendar that sold out suspiciously early. The winter third shows up as empty midweeks you blamed on the weather. The events third leaves no trace at all, because those nights sold anyway, at the wrong price.
Let us be your Host B
That daily, mechanical attention is exactly what RateVeld does for members. It reads the comparable listings and the hotel rates in your own suburb every day, knows both the national and the independent-school calendars, keeps a hand-curated Cape Town events calendar wired into the curve so a race Saturday reprices on announcement day, and prices every night of your calendar for the next six months, in rand. It is calibrated on South African market data rather than a northern-hemisphere pattern, shows the reasoning behind every rate, and never recommends a single rand below the floor you set yourself. An event week priced this way can more than double an ordinary one, which is the events third of the gap, arriving as money; how much depends on your suburb, your season and how you price today.
Plans start at R333 a month, in rand, a good deal less than the flat next door is quietly earning over yours. Pick one this week, and stop finding out in January which host you were.