With the poll thing, since you mentioned you worked in market research which is certainly more experience than I have, I had a question. I was taught in school that you generally wanna collect demographic information to do comparative analysis (does x age group have a different trend than y age group? etc) or to weight responses against your data frame, so you can do non-response analysis.
given we don't have a data frame for tumblr users (I think. Correct me if I'm wrong), and that tumblr polls don't let you see what was selected on an individual level - only cumulative counts, what is the benefit for including that?
(note, i'm just curious, def not trying to throw shade or anything, so if the answer is as simple as "eh, wasn't paid to do complicated shit tm, so I didn't" or smth similar, that is super fair.
So first, I need to clarify that when I say I worked in market research, I very specifically mean I was the person asking what kind of paper towels you buy. (Okay, I never actually did a survey on paper towels, but I did do one on toilet paper and the number of men whose immediate response was "uhhhhhh.....let me get my wife" was both entertaining and horrible.) So I learned how to construct a questionnaire based on seeing them in action, not from school. With that said: demographic data can serve two purposes. One is to make sure you're hitting all your groups--for example, let's say you're running a survey to test responses to a new toy, and you include four separate age groups under the age of 18: 4-7, 8-10, 11-13, 14-17. You get 460 in the first group, 520 in the second group, 78 in the third group, and 455 in the fourth group. It's going to appear that ages 11-13 are less interested in the toy than the other three groups even if that group demonstrates more interest, because there aren't enough of them. Plus, you've got another problem--let's say this is a science toy, and your 78 tweens all got here via a flyer at a science fair. Cool! The per capita shows this toy is heavily favored by tweens! Let's market it to....ah. If you'd had another 400 people in that group you'd see it might be favored, but not "heavily" so. You need that larger group to ensure you've got a good sample size. The other reason to collect demographic data--and this is why I did it on the poll set--is to determine who your audience is in the first place. For example, if you take a look at the age responses on the poll set, you'll see the vast majority of us fall between 18 and 45. Now let's say this poll is being used by Tumblr for advertising purposes. Which of the following ads would be most likely to get play among that age range: an ad for a wealth management company, an ad for trendy teen clothing, or an ad for something that helps you organize your kitchen? You've got a bunch of broke young-and-mid adults. Most of them probably aren't wearing teen clothing anymore and most of them are in the wrong age range to be purchasing it for their own children. Most of them are too broke to be worrying about wealth management, and a lot of them are young enough they're more worried about rent than retirement. But all of us have to eat and a lot of us are in the age range where we're cooking for ourselves.
The reason I felt demographic data would be useful here is because there's a very noted trend where younger users (super-early 20s and younger) trend toward video, while older users (definitely Millennials and Gen X, some early Gen X) trend toward text. This is a trend, not an absolute, but it's good information. If most of your users aren't creating video-based content, then optimizing tools for text and still images should be a higher priority. You want to have a good, stable platform before you go making changes to it. It's also true that the length of time spent on Tumblr is important because this site came into being right around the end of first-wave social media, or as I put it in the poll "pre-social media", dominance. That's not to say those places aren't still around--they are--but they've been overtaken. That isn't always by the choice of the users, though. If I had my druthers LiveJournal would still be kicking hard. And so: people who've been on this platform since those days, who came here when it was substantially more similar to places like LJ and early MySpace (which I think a lot of younger users would be surprised used to be a lot like a proto-Tumblr), may also trend toward preferring older features that don't rely on video or algorithm. And indeed, the answers to the "preferring older platforms to newer ones" question seems to bear that out.
The last question, admittedly, was mostly for fun. But as someone who's been here long enough my Tumblr Experience is old enough for high school (this blog dates to 2014, but I had an RP blog here for a year or two before that), I feel like two things are true: one, you have to let the people have some fun and I knew people would be getting antsy without their vanilla extract. And two, if @photomatt and @staff think that experience is actually worth hiring on, yes, I'd totally do it.
