Reimagining Decision Confidence in Phia's Shopping Experience

The Problem

What would it take to never regret an online purchase again?

Flew from SF → NY to pitch my solution answering this question in front of the Phia team.

24 hours. Selected from 50+ submissions. 1 challenge: design an end-to-end feature that solves one high-friction moment in online shopping.

Context

85% of online shoppers regret an impulse purchase according to Investopedia.

Shopping is currently a broken process. Drowning in analysis-paralysis. Second-guessing decisions.

Introducing Circle Vote

Before you buy something, you ask your inner circle whether you should buy it.

One tap. Trusted people. Real-time votes. Opinions from humans who actually care if you regret it.

User Understanding

Meet Carmen, the “Overthinker”

28, works in tech sales, shops online 3-4x/week

• Has 50+ items saved in the carts across multiple sites
• Screenshots products to friends and hates waiting an unknown amount of time for them to respond

Meet Emma, the “Go-to Advisor”

29, works in fashion, known for good taste in friend group

• Gets 6-10 “Do you think I should buy this?” texts per week
• Loves to help but feels overwhelmed with context-switching between messages in different places

PHASE ONE: The Extension

I started with tackling the core design challenge: How to make asking for help effortless?

Minimizing taps was the main goal. Every extra step in a moment of impulse is a potential exit ramp. I had to make it feel like a reflex, not an unfamiliar feature.

Decision 1

Product context always visible

Product image, price, and brand stay in view entire time. No switching screens, no losing context.

Decision 2

Make personal context optional

Sender's note gives the circle better signal to vote. But requiring it can kill the momentum of the moment.

Decision 3

Default everyone in, allow opt out

Your full circle is toggled on by default. Removing someone is a lighter action than adding them. Fewer taps, faster send.

Decision 4

Let sender set the urgency

Not every purchase has the same timeline. Sender controls the pressure so that the circle knows exactly how fast their opinion is needed.

Once the vote is live, the sender watches their circle respond in real time.

Comments roll in alongside a live consensus bar showing where the circle is leaning.

6 hrs remaining

PHASE TWO: The App

During the feedback session, the Phia team asked how would this translate to the app.

This was not just a resize challenge. This opened the context wider, to the device where the circle usually lives. And begs the question of how do people respond to a vote.

The Built-In Growth Loop

Every time a user sends a vote request, they're potentially funneling people into the Phia ecosystem who may have never heard of it.

If someone in your circle doesn't have Phia, they get a text. The same way you'd hear from a friend. The vote only means something if the voter is invested. Filter for people who actually care and get a real user out of it.

Beyond the Competition

Given more time, I'd conduct deeper research and continue building out the product.

I identified five metrics to measure whether Circle Vote actually works.

The two flows I'd prioritize next: the circle creation process and a personal votes dashboard where users can track active votes and see how their circle responded.