Shopié
Reimagining social commerce through AI-powered discovery, bargaining, and trust.
Lead Product Designer
2 months
Overview
Most shopping apps get deleted after the first use. I designed one that turns price negotiation into a core retention loop.
Digital commerce has perfected logistics, but it completely erased one of the most fundamental human shopping behaviors: haggling. Fixed pricing works for mass retail, but in social commerce, buyers want to negotiate, and sellers want to clear inventory.
When platforms force rigid pricing, users don't stop bargaining—they just leave the app to haggle in Instagram DMs or WhatsApp.
I was brought on as Lead Product Designer to build a digital negotiation system that digitizes real-world bargaining mechanics without the friction, slowness, or awkwardness of manual messaging.
Contributions
I led the entire design process, from research and strategy to testing and refinement.
Responsibilities:
Lead Product Designer: Solo designer overseeing UX research, interaction design, micro-copy, and high-fidelity prototyping over 2 months.
Core Focus: Designing an asynchronous AI-mediated negotiation system, offer state models, and trust mechanics.
Problem
Fixed pricing ignores how people naturally shop, but manual bargaining in DMs breaks down at scale.
When I looked at how users buy through social channels, the friction wasn't in finding the product—it was the painful back-and-forth required to settle on a price.
Bargaining is emotional, but fixed pricing is rigid
Users feel a psychological high when they negotiate a deal. Fixed-price apps kill that engagement, leaving users feeling like they might be overpaying.
"I always offer less. It's part of shopping."
Manual haggling in DMs creates high friction and ghosting
When buyers try to negotiate manually, they get stuck in slow DM threads. Sellers miss messages, buyers lose interest, and lowball offers create interpersonal tension.
"I'll ask 'How much last?' and wait six hours for a reply. By then, I don't even want it."
The trust gap ruins the deal at the finish line
Even when a buyer and seller agree on a price in DMs, there is no system to lock in that price. Buyers worry about getting scammed, and sellers worry about buyers flaking.
"We agree on a price, but then they ask me to send money to a random bank account."
Process
To digitize negotiation, I had to break down the psychological rules of a real-world deal.
I conducted interviews with buyers and informal sellers to map out the exact micro-interactions that happen during a physical market negotiation.
Stage 1: Deconstructing the Negotiation Loop
Four critical insights emerged regarding how people actually haggle:
The Anchor Price
Buyers need a starting reference point, but they rarely want to make the first move if they don't know the seller's floor.
The Counter-Offer Threshold
If a counter-offer takes longer than a few minutes, the emotional impulse to buy evaporates.
Lowball Protection
Sellers hate insulting offers, but they don't want to turn away serious buyers who are close to their target margin.
Instant Commitment
The second an offer is accepted, the deal must be locked immediately before buyer regret sets in.
Stage 2: Mapping the Interaction States
Unlike a simple "Add to Cart" button, bargaining requires a multi-state logic system. I mapped out four primary states for the core mechanism:
State A:
Initial Pitch (Buyer sets offer within allowable guardrails)
State B:
Instant Mediation (AI evaluates seller rules in under 3 seconds)
State C
Counter or Accept (System presents a fair middle ground or closes)
State D:
Price Lock & Checkout (Immediate transition to payment)
Solution
Replacing awkward messaging threads with a structured, instant negotiation engine.
Instead of building a messaging chat box, I designed a specialized Haggle Interface that turns negotiation into a fast, game-like interaction while keeping human psychology at the center.
Designs
The AI Bargaining Assistant
People want a better deal, but many hesitate because they don’t know what offer is reasonable or how to negotiate without offending the seller.
Instead of forcing users to type an offer or start an uncomfortable negotiation, I designed an AI-powered bargaining assistant that handles the conversation on their behalf.
The AI acts as a smart shopping companion, understanding the product value, seller pricing, and previous deal patterns to suggest a fair offer and guide the negotiation naturally.
By turning bargaining into a simple conversation, users can confidently negotiate in seconds without the pressure of choosing the “right” words or making an unrealistic offer.
2. AI-Powered Instant Mediation
Negotiation loses its excitement the moment you have to wait hours for a reply."
Sellers pre-configure their minimum price floor and margin rules behind the scenes. When a buyer submits an offer, the AI co-pilot evaluates it instantly on behalf of the seller.
Instant Accept: If the offer meets the seller's threshold, the deal is accepted immediately.
Smart Counter-Offer: If the offer is too low, the AI generates a firm counter-offer midway between the buyer's bid and the seller's list price, keeping the deal alive.
No Waiting: Cuts negotiation time from 6+ hours in DMs to under 10 seconds.
3. The Deal Lock & Instant Checkout
An accepted offer is useless if the buyer walks away five minutes later."
The second a price is agreed upon (either through instant acceptance or a accepted counter-offer), the UI instantly transforms into a time-bound checkout screen.
Price Lock Guarantee: The negotiated price is locked for 15 minutes, creating subtle urgency.
Escrow Safety: Funds are held in escrow, reassuring the buyer that money won't be released until the order is shipped.
Zero Disruption: Eliminates the need to copy-paste payment links or switch to third-party banking apps.
More Screens
Impact
Turning negotiation into the core engine for engagement and conversion.
3x Faster Deal Resolution
Reduced average time-to-agreed-price from 4.5 hours (via social DMs) to under 15 seconds through automated mediation.
Drastic Reduction in Abandoned Intent
Cart drop-off plummeted because buyers felt an immediate sense of ownership over a price they personally negotiated.
High Seller Retention
Sellers reported zero time wasted answering repetitive "How much last?" messages, while maintaining their profit margins through pre-set AI guardrails.
Learnings
The best feature designs don't introduce new habits, they digitize existing human instincts.
Product design is behavioral psychology first, UI second.
The hardest part of this project wasn't making screens look pretty; it was designing an interaction model that removed social anxiety from bargaining.
Automation must preserve human feel.
The AI couldn't feel like a cold bot declining a transaction. Framing counter-offers with language like "The seller can do [Price] if you lock it in now" preserved the feeling of a personal deal.
More projects
Check out some of my favorite & most recent projects.
