Tag Archives: YouTube Analytics

QuickNote: Dealing with Demographics

8 Dec

Today I got my first good look at YouTube Analytics for the webseries.  There’s all sorts of fun statistics to work through (NERD!), but I’m trying to focus specifically on the aspects that are the most useful to the client: who is watching, and how they’re watching.  (Monetization falls outside the scope of this project.)

The issue I’m puzzling out today is how to deal graphically with some unexpected demographics.  Before starting the project, I expected that a parody show about superheroes that runs on YouTube would primarily attract males, age 25-45ish (Gen X and Y).  Mom and Pop are also entrenched in geek culture (“nerd famous”?), so they seemed likely to see some crossover from fans of comics, graphic novels, manga and video games.  Kind of the “Wreck-It Ralph” crowd.

The gender demographics aren’t too surprising: 65% male viewers to 35% female viewers.  (Need to figure out whether that’s lifetime or per video, and what percentage of those are unique vs. repeats.)  But within those gender constraints, the age graphs look radically different.  I.e., the male graph follows a bell curve, peaking around 35-44, while the female viewership peaks at 18-24 and then drops off a cliff (f(x) = 1/x where x>0).  The difference is pretty stark.  However, if you average the values by age group, weighing them proportionate to the total percentage of the population, the curve becomes bell-like again.

There isn’t time on this project to create formal user personae or conduct tests.  But with such a stark difference in viewing populations, I need to figure out a meaningful way to describe those behaviors.  The clients might want to factor them in if they decide to include advertising, for example.