Using the following dataset about hotel performance, we will examine price and occupancy of Brussels hotels by region and see whether they could be related in some way. The data refers to the period between January 2013 and July 2017, so anything you see here will be valid in that time interval.
We will start with the average price of hotel rooms over time, broken down by region.
We see that Quartier Européen tends to have the highest average price throughout the year, while Midi-Lemonnier the lowest. More precisely:
|Region||Avg. price (euro)|
|Grand Place & environs||109.72|
|Autour du Ring||89.45|
|Grand Place & environs||0.7454|
|Autour du Ring||0.6362|
From the first table, we can see that by preferring Quartier Européen over Midi-Lemonnier, tourists can expect to pay ≈47% more, on average. We also see that the occupancy of the hotels in the second region is on average 5% higher than in the former.
But the figure also reveals an interesting periodic pattern: the prices of hotel rooms seem to vary over time in the range between 70 and 150 euro. There are periodic lows which might correspond to certain times of the year, so it is interesting to observe whether there is some variability in the average monthly prices.
In August and partially in July we see the lowest average prices of around 80 euro, while in March, April and September, we see the highest—around 120 euro. The lows that we have seen then in the first figure are likely to be related with July and August. What is also intesting is that during these months, the hotels in all regions reduce their prices a lot, so that the price difference (spread) between different regions becomes minimal. It is then less important to have a preference for a particular region as the prices become very similar.
Next we see a figure showing how the occupancy of the hotels has varied over time.
We see that the hotels were most often occupied at 50-80% (or slightly more). They don't seem to fully utilize their existing capacity. Further reducing prices could help with this, but it is questionable whether it would be economical to do so. Further, we see that after a more recent drop in the occupancy level, the hotel owners has once again started to register more people.
We can also plot average occupancy vs. average price by region to see how the regions position themselves.
In the case of Midi-Lemonnier we see that higher occupancy doesn't necessarily mean higher hotel room prices.
We could also plot occupancy vs. price for each region, which would give us the following noisy figure.
If we look at the black line, which resembles the average for all regions, we can see that as the occupancy grows, the price tends to increase slightly. If we were to draw a least squares line passing as close as possible to all data points on the black multi-segment line, we would get a new line with some slope. We see that an increase of the occupancy between 50 and 85%, tends to increase the average price from 100 to 115-120 euro. For individual regions—like Quartier Européen, for instance,—the slope could be even more pronounced.