Inspiration
We were inspired by the story of Target building a model that could predict which customers were pregnant from what they bought, even estimating their due dates. Target then sent those customers baby-product coupons, deliberately mixed in with unrelated ads so they wouldn’t feel watched. What stuck with us wasn’t just that a store could figure out something that personal from a shopping cart. It was that the company actively hid what it knew. If stores are already drawing conclusions about us and designing around our not noticing, then the prices and offers we see are shaped by inferences we never get to look at.
What it does
Peekaboo appears inside any shopping app and shows what most people pay for an item next to the price you’re being charged. It explains the difference by naming what the store has inferred about you, like late-night shopping or viewing an item several times, and how much each one adds to your price.
How we built it
We started with a standalone app, but asking people to keep switching to a separate app to track their own shopping didn’t make sense. So we moved to a layer that appears inside any shopping app. We combined two ideas: a “horoscope” that describes the shopper profile a store has built about you, and a “price reveal” that shows what you pay compared with most people and how much each behaviour adds. We designed and prototyped Peekaboo in Figma, with a witch mascot, a short onboarding flow and alert settings. The prototype covers two shopping flows that show a markup and a discount, with a swipeable carousel of traits. To answer “who would actually build this?”, we framed it as speculative design: imagine a transparency law required platforms to show you their inferences.
Challenges we ran into
The hardest part was keeping it believable while being honest that much of it couldn’t exist today. Platforms don’t share how they set prices, phones don’t let one app appear inside another, and real pricing models are far messier than “+$8 for shopping at 1 a.m.” We also had to balance simplicity against accuracy: archetypes make the data relatable, but they can oversimplify people
Accomplishments that we're proud of
A clear concept: we got the idea down to one sentence, “Imagine a transparency law required platforms to show you their inferences. This is what it would look like.” It answers “who would build this?” before anyone asks. Pairing a profile with a price: the horoscope explains why you’re seeing a price, and the price reveal shows what it costs you. Linking each behaviour to a dollar amount makes an abstract practice feel concrete and personal. An unserious tone: plays into the uncomfortable nature of this data. We intend the unseriousness to be a critique of how absurd it is that these are being tracked in the first place!
Link to our slides: https://www.figma.com/deck/JsTv85TMcKdRlh46mGWi4J
Built With
- claude
- figjam
- figma
- imovie
