I assume you all sometimes look at the demographics of your listeners, such as their age, right? In my opinion, the graph that Spotify uses for this gives a quite distorted picture; the age groups used are not all the same size. One group (for example those aged 23 to 27) covers 5 years, while another (those aged 45 to 59) covers no less than 15 years. That makes a comparison difficult. As an example, see the graph of my listeners age (over the past 28 days) below.
Now I like statistics, so I did some simple math to see in which age group people actually listen to my music the most. And guess what? By far the majority (see below) of my listeners are between the ages of 18 and 34, with an emphasis on the group aged 23 to 27. Quite surprising, I thought 🙂
oh yeah you already said it @Liam Phan
hey Geert, I did not notice that unequal age division yet. But for me it's quiet simple in the reason: when the amount of spotify users around 23-27/28-34 are simliar to 45-59 (which is what I believe) it makes sense to make this devision. That's my presumption behind it: The number of people behind each bar, not the difference in age. Anyway, I don't find it the most important stats out there (and I love stats) "gender and age" you can question a lot of those in addition. Not everybody is entering his real birthday to services.
I do look at those stats from time to time but simply from curiosity... i've never truly needed this kind of information from my listeners so i don't mind it not being 100% accurate. I docare about the location where my audience is listening from though! but as far as male/female and their age, atleast for now, i dont care much!
I guess I might give you a simple reason: distribution.
Using equal intervals might not be very interesting for spotify to use it that way, because most ages are not randomly distributed, and you always end up with more people in their mid 30 than older people, so what most social medias do is using the basic demographics distribution, so that ages are divided in proportion of the normal distribution using quartiles. I might be wrong about spotify, but I know most ages group are done that way and also based on most legal laws, that’s why we start at 18 years old and stop at 60+, such that retired people are most accounted in that category, and 18 years old, are mostly not student, and need to comply into the working phase (so no premium at student price)
But still, what you want to do is not a interval category, but instead just plotting by individual age the distribution of listeners, use normal distribution / box plot, and you will actually find your median or average listeners in there