Sketching activity
For this activity, submissions are individual, but you’re encouraged to talk to the people around you as part of the activity (more on that below). You’ll turn in photos of your sketches along with a short write-up through Canvas. Phone photos are fine; legibility matters more than artistic quality.
Grading note
This is an in-class participation activity, so on-task responses will be graded for completion rather than “correctness”. Messy, discarded sketches are encouraged and expected — that’s the whole point.
Scenario
Imagine you have access to a year’s worth of a Spotify user’s listening history. It can be your own, a friend’s, or a hypothetical user you invent. Think about what fields plausibly exist in that data: songs, artists, albums, genres, timestamps, play counts, skip rates, device, time of day, and so on.
Your goal is to design a single static visualization that reveals something interesting about this person’s listening habits. Think “one Wrapped slide”, not the full Wrapped presentation. You will not write any code today — everything is pen and paper.
(See here for more information about Spotify Wrapped.)
Two things to do (not in order)
The activity has two strands, and the whole point of today is that you should be spiraling between them, not marching through them in order. You’ll start with a vague idea, sketch something, realise your audience is wrong, rewrite your framing, sketch again, realise the encoding can be improved, and so on. That’s the exercise.
Write/mark things as you go — even scribbled margin notes on your sketches are fine.
Framing (who, what, why)
Using the outer levels of the nested model, figure out:
- Who is this for? The listener themselves? A friend they’re sharing with? An Instagram audience? What does that person already know, and what do they care about?
- What data are you assuming you have? Be specific about fields and their types.
- Why is the viewer looking at this? What’s the specific insight? A good answer reads like “I want the viewer to notice that…”.
Expect your answers here to change once you start drawing. That’s fine — cross things out, add new answers, let it be messy.
Sketch, and throw most of them away
Make several sketches and expect to discard most of them. Aim for at least 4–5 distinct attempts. They should be fast and rough — if you’re spending more than a few minutes on one, you’re overcommitting.
For each sketch, jot down a few words about encoding choices: what are the marks, what’s on each axis, what does colour mean? Note what you liked and what felt off.
Talk to your neighbours
You should pause and talk to the people sitting around you (and they should do the same to you!). This’ll be useful for inspiration.
For example:
- Early (once you have rough framing): describe your listener and your “why” in one sentence to a neighbour. Ask them what they’d want to see, and what data they’d look for. Steal ideas shamelessly.
- Mid-way (once you have a few sketches): show somebody your scribbles. Feedback along different levels of the nested model are going to be very useful here.
Before you submit: pick one and say why
Circle (or otherwise clearly mark) the sketch you’d take forward to a higher-fidelity prototype. In a short paragraph, answer:
- Why this one and not the others?
- What did the discarded sketches help you figure out? (Even bad ideas usually teach you something.)
Submission
Submit to Canvas: photos of all your sketches (including the discarded ones), your who/what/why framing, and your short reflection paragraph. If you prefer to draft text in a Google Doc first and paste in, that’s fine — just make sure the images come through.