July is not January.

In that same SiteMinder set, July’s average daily rate was $269.08. January’s was $217.15. July and August together were 20.74 percent of check-ins. January was 6.05 percent. These are the same hotels and the same year. They are not the same price, and they are not the same volume of arrivals.

The two facts should be read together. In summer, far more people arrive and the night costs more. In January, fewer people arrive and the night costs less. The year has a shape. Someone who has run a hotel for twelve months already knows it, before seeing it written down. The report does not discover it. It measures it.

The average for the year fell 2.89 percent, from $254.39 to $247.03. It is a real figure, and it is worth having, because it says something about the year as a whole. It says nothing about one month. 247.03 falls between 269.08 and 217.15. It is not July and it is not January. Nobody paid that average in July for a reason connected to January. Two different seasons have been put into one number.

It also does not say how the drop happened. A yearly average that falls 2.89 percent may be July, January, both, or other months the figure does not separate. To know which, the months have to be looked at. The average does not do that.

The reason beside a July rate cannot be the reason beside a January rate. Arrivals, demand, and the pressure from one week to the next all change. If the explanation talks about the year, it is talking about a line that no month followed. A useful explanation names the month, just as the previous note’s explanation names the day.

Which is why it makes little sense to explain a July rate with the yearly average. In that set, July and January shared neither the price nor the volume of arrivals, and the yearly average describes neither. Compare the month with the same month, not with the whole year. If you only have the yearly figure, you still do not know what each season did.