All work

Tandem

Concept

A group-trip concept that gathers shared dates and private budgets, compares a few weekends on the same facts, and records a plan only when every person says yes.

Role
Product strategy, UX/UI and prototyping
Timeline
2026, self-initiated.
Year
2026
See it
Try the prototype
Tools
Figma · HTML/CSS prototype · Claude
Type
Mobile app · Prototype

00 · TL;DR

The thirty-second version.

Six lines on Tandem. The chapters below show the work behind each one.

Role
Everything: product strategy, UX and UI, the prototype.
Team
Solo, with Claude as a build partner for the browser prototype.
Timeline
2026, self-initiated.
Scope
Ten app screens, a prototype with three testable states, and a short landing page.
What I decided
Find what works for everyone before asking what anyone prefers. Keep budgets private. Silence never counts as yes.
Outcome
A clickable prototype anyone can try. No users and no metrics: it is a concept, and the page says so.

Chapter 01 of 06 · Overview

A weekend everyone can say yes to

A group trip can start with one message and stall in a hundred replies. Tandem explores a calmer path: shared dates and private budgets first, then a small shortlist compared on the same facts, then one explicit decision.

FIG. 002Start, compare, agree: the whole idea in three screens.

Chapter 02 of 06 · Sequence

Find what works, then choose

An early version asked which destination was most popular. That could produce a winner while hiding a date or cost conflict. I changed the order: find what works for everyone, then express a preference. The budget is a private limit, so nobody has to negotiate their number in front of the group.

FIG. 003Dates first, then a private limit, then a shortlist — the order that keeps a conflict from hiding behind a favourite.

Chapter 03 of 06 · Agreement

Liking is not agreeing

Liking an option and committing to it are different actions. Tandem gives the group a review step with the weekend, dates and estimated cost in one place, and the plan is agreed only when each person responds. Silence does not count.

FIG. 005Review, respond, then plan. The plan exists only after every yes.

Chapter 04 of 06 · Edge cases

When nothing fits, say so

If no option fits the group, the no-match state keeps the constraint visible and offers a way back to dates or budget. If the estimate changes materially, earlier agreement clears and everyone is asked again — protecting the meaning of the original yes.

FIG. 007When nothing fits, say which constraint failed. When the price moves, ask again.

Chapter 05 of 06 · Prototype

Clickable, not just drawn

The prototype runs in the browser at phone size, with the default path, a no-match state and a changed-estimate state all testable. A short landing page explains the idea in four steps — Align, Compare, Agree, Go — before anyone opens it.

FIG. 009The landing page at 390 px.

Chapter 06 of 06 · Final work

Every screen, as designed.

10 screens from the finished work. Open any of them full size; arrow keys step through.

  • Landing page
  • Landing — hero
  • Landing — four steps
  • Landing — private budget
  • Prototype in the browser

What shipped

  • A working prototype with default, no-match and changed-estimate states.
  • An agreement model where silence never counts as yes.