Like many new (or soon-to-be-new) authors dipping their toes into social media, I’ve been trying to get a handle on not only how to better manage the wild chaotic beast that is Twitter, but also how to understand it. Eva Gantz is the go-to expert on social media for authors, and following her advice I checked out Twitter Analytics, which I discovered has this cool little feature for breaking down the demographics of your followers.
Neat! So now I know that 55% of my followers are in the US (no surprise there, because I’m in the US, I suppose), 17% in the UK, etc. I know that 88% are interested in “Romance (books and literature),” which makes sense, because I’m trying to be a romance author — that’s probably the most useful statistic, actually.
And I know that 51% are male and 49% female.
Huh? How does Twitter know that? I wracked my brain trying to remember if Twitter ever asked me my gender when I signed up, and no, it didn’t. I checked my profile, and nowhere is there anything that says “female” (or any other gender!). So, if Twitter doesn’t know my gender, how can it know the gender of my followers?
The answer is, apparently, big data and social algorithms. It is all in your word choice, word frequency, word proximity, etc, apparently. Twitter (and Facebook too, for that matter) has been hiring some of the world’s foremost social statisticians and researchers, because they want to know this sort of thing about us (without asking, apparently).
I guess it’s a little creepy and a little fascinating at the same time.
Glenn Fleishman wrote a very interesting article for BoingBoing about it a couple months ago, which I came across in my search for the answer. It seems he had the same head-scratching moment I did when he looked at that graph. Oddly, it appears for most people their followers skew far more to the male side, regardless of their own gender or what they tweet, whereas mine are nearly evenly matched. I don’t know that that’s good or bad or indifferent — I’m very happy if men find what I have to write as interesting as women do, and in fact I do have some men who regularly read my blog, and I love them for it (you know who you are, and you’re all wonderful)!
Have you looked at your own Twitter Analytics results recently? What is your thinking about computers determining your gender based on how you write?
There’s a lot more to the analytics that I haven’t even begun to understand yet. There’s a lot to Twitter I don’t understand yet.