An average is a fair summary only when the things you average are alike. Property sales in one area rarely are. The gap between an area number and a building number can change which flat you buy and what you pay for it.
How one average hides two different markets
For example, take an area with two buildings and some invented, round numbers. The first building is an old block with modest flats. The second is a new tower with a pool, a gym and wide views. In one month, ten flats sell. Eight sell in the old block at AED 800 per square foot. Two sell in the new tower at AED 2,000 per square foot.
The average for the area is AED 1,040 per square foot. No flat in either building sold at that price. A reader who sees “AED 1,040 for the area” may judge old-block flats cheap and new-tower flats expensive, when each building is simply priced as what it is.
Now change the mix and nothing else. Next month, five flats sell in each building at the same prices as before. The area average rises to AED 1,400 per square foot, an increase of more than a third. Neither building became more expensive. The average moved because the mix of sales moved.
What to ask for instead
Platforms that start from registered sales can answer building-level questions. dataHabibi, for one, builds its figures from transactions registered with the Dubai Land Department and gives each building its own price history.
When someone hands you an area number, ask five questions:
· How many sales sit behind it? Ten sales and ten thousand sales are different kinds of evidence.
· What period does it cover?
· Which buildings are in it, and how do they differ in age, size and finish?
· Where do the cheapest and the priciest sales sit, apart from the middle ones?
· What is the figure for one building and one flat type?
The range question deserves a note. A median, which is the middle sale, helps a little but not enough. In the example, the median for the first month is AED 800, which hides the new tower altogether. A range of AED 800 to AED 2,000 shows at once that the area holds two markets.
The last question matters most. A figure for a single building and a single flat type compares like with like, so a change in it means something.
A ten-minute test
If you can get a list of recent sales, test any area number yourself. Group the sales by building. Work out the middle price per square foot for each group. Then compare each building’s figure with the area figure.
If the buildings sit close to the area figure, the average is a fair guide. If they sit far above and below it, as in the example, the average describes nobody. There the old block sits at AED 800 and the new tower at AED 2,000, and both are far from AED 1,040.
Do not stop at one month. Run the test on a second month, and see whether the buildings kept their places while the area figure jumped about.
How building numbers change a decision
With only the area figure, you decide on the area. With building figures, you decide on a building.
Return to the example. A buyer who wants the new tower can see that flats there sell near AED 2,000 per square foot, and can judge an asking price of AED 2,300 against that. A buyer of an old-block flat can see that sales cluster near AED 800, and can question an asking price of AED 1,100.
A building’s own history also shows whether it moved with its area or against it. If the area average climbs, the history tells you whether your target rose too, or whether the average only reflects a change in which flats happened to sell. That is the difference between a trend and an accident of the mix. The same problem affects rents and yields, so ask for those by building too.
Area figures are not wrong. They show which part of the city is moving. Building figures help with the next question: what to pay for this flat. For that, datahabibi.ae shows the figures one tower at a time.
