# FitMatch — Cross-Brand Fit Prediction

A research-backed product management case study applying evidence-first problem discovery, competitive validation, and honest AI-scoping to the cross-brand apparel sizing problem.

---

## Thesis

Existing sizing tools fail people who want to try a new brand because they solve the problem *within* a brand, not *across* brands — they need body scans or brand-side data the shopper (and an independent builder) doesn't have.

This case study argues the missing piece isn't better prediction technology. It's a **user-owned fit profile** — a sizing signal that travels with the shopper instead of living inside one brand's system.

FitMatch does not ask "what are your measurements?" It asks "how has what you already own fit you?" — and infers the rest.

---

## Core Insight

Most sizing tools intervene *after* the wrong size has already shipped — at the return, when the cost is already sunk.

This case study identifies the size selector on an unfamiliar brand's page as the primary high-leverage intervention point: the last moment where a prediction can still change the outcome, before a guess turns into a return.

FitMatch intervenes at that moment. Everything before and after it is out of scope for v1.

---

## Case Study Structure

| Phase | Focus | Status |
|---|---|---|
| 01 | Problem Research | ✅ Complete |
| 02 | User Research | ✅ Complete |
| 03 | Solution Design | ⏳In Progress |
| 04 | Ethics & Metrics | ⏳ Pending |
| 05 | Final Documentation | ⏳ Pending |
| — | Working v1 prototype | ⏳ Pending |
| — | Verified size-chart data | ⏳ Pending |
| — | Real user testing | ⏳ Pending |

---

## Phase 01 — Problem Research (Complete)

### market-analysis.md
The scale of the problem and why it's still open. Key findings: 53% of retailers cite size/fit as the top return reason; $38B in annual returned apparel; despite heavy industry investment (85% adopting virtual try-on), only 1 in 4 retailers offer cross-brand fit guidance.

### competitor-audit.md
Why the original wardrobe/styling app direction was killed, and why cross-brand sizing was selected instead. Key finding: the wardrobe app category is crowded and the gaps there are execution gaps, not opportunity gaps. Sizing, by contrast, is well-attempted but still structurally unsolved across brands.

### phase1-synthesis.md
The one-page argument: the industry has built size intelligence within brands, not across them. That's the wedge.

---

## Phase 02 — User Research (Complete)

### personas.md
Primary persona: an online shopper whose hesitation isn't about price or style — it's specifically "will this fit," surfacing at the moment of selecting a size on an unfamiliar brand.

### journey-map.md
Five-moment map of a cross-brand purchase decision. Critical finding: the coping response (over-ordering, cart abandonment) is a diagnostic signal of the problem's cost — the fix belongs earlier, at the size-selection moment itself.

### opportunity-spaces.md
Identifies the size selector as the primary intervention point, and makes the case for why AI (inference from subjective fit feedback) is justified there over a static lookup or hand-coded rules.

---

## Phases 03–05

Still working on the product and the phases.

---

## Why This Project

Most product case studies show either the research or the build, rarely both connected end to end. This one traces every design choice back to a research finding or a killed alternative — including the parts that didn't survive.

The persona isn't invented. The intervention point wasn't assumed. The problem wasn't decided in advance — it emerged from discovery, and a stronger direction (this one) only got selected after a weaker one (the wardrobe app) was tested against evidence and killed.

---

## Author

Niharika Chauhan
B.Tech CS + AI/ML, VIT Bhopal