Product & UX Case Study
Pattarai
பட்டறை · Tamil for “workshop / forge”
Building trust into how Tamil Nadu hires contractors — a two-sided marketplace for full construction projects, not one-off repairs.
Role
Product Strategy & UX Lead
Type
0 → 1 product concept
Market
Tamil Nadu construction
Status
Validation phase
ROLE
Product Strategy & UX Lead — problem framing, research, MVP definition, GTM
FOCUS
Marketplace strategy · trust & verification design · adoption psychology
TYPE
0 → 1 product concept (self-directed)
MARKET
Tamil Nadu construction & renovation sector
STATUS
Validation phase — pre-build (no-code Phase 0)
PRINCIPLE
Evidence first, code second, capital third
Overview
Pattarai reimagines how households and businesses across Tamil Nadu hire contractors and skilled tradespeople — not for quick single-visit fixes, but for the harder problem the big apps ignore: full construction and renovation projects that run for weeks or months, coordinate several trades, and depend entirely on trust between people who have never met.
I led product strategy end to end — framing the problem, running early field research, defining the MVP, and shaping a go-to-market approach built on one conviction: the real problem here isn’t discovery, it’s trust. A better search box doesn’t beat a cousin’s recommendation. So instead of competing with word-of-mouth, the product is designed to digitize it.
94k+
construction firms in TN, mostly offline
~79%
of clients search online before hiring
20–25%
typical cost / time overrun on projects
₹1.5L–15L+
value of a single project
Figures drawn from Tamil Nadu construction-sector and digital-adoption research (2026). Full-project construction remains largely untouched by dedicated platforms, even as billions flow into quick-fix home services.
Context & why it matters
Anyone who has renovated a house, built on their own land, or fitted out a shop in Tamil Nadu hits the same wall: there is no reliable way to find and compare good contractors unless someone in your circle already knows one. The entire multi-week, multi-trade construction economy still runs on personal referral.
What made this project unusual is where it started. The founding team has a built-in, trusted supply network — a working-contractor father and an extended family spanning plumbing, electrical, welding and carpentry, plus a civil-engineering sibling who is qualified yet struggles to find consistent project work. That gave the venture something most marketplaces never have on day one: a real anchor supply to launch with, instead of a cold-start problem. My job was turning that lived, on-the-ground advantage into a defensible product strategy.
The problem — two broken sides
For clients — trust with no infrastructure
• No way to verify a stranger’s track record before money changes hands — worse for first-time builders, remote NRIs, and anyone outside the big cities.
• A referral pool that shrinks fast in Tier-2 and Tier-3 towns, where most construction actually happens.
• Little recourse when a contractor underdelivers, because there was never a formal agreement, review, or record.
For professionals — skill without reach
• Fifteen years of good work exists only in the memories of past clients — no portable, verifiable track record a new client can check.
• Idle periods even when the work is excellent, because reach ends at the edge of one neighbourhood.
• Qualified newcomers — a fresh civil engineer, say — can’t get started at all, because the world runs on “who you already know.”
Why the existing apps don’t solve it
India’s on-demand home-services category matured fast — Urban Company, NoBroker and others let you book a plumber in minutes. But that entire category is built for short, low-value, single-visit jobs. None of it is designed for a multi-week renovation or a new build that needs a civil engineer, mason, electrician and carpenter working in sequence over months. That gap — full-project discovery and trust infrastructure — is the whole opportunity.
Research & the insight that reframed everything
I grounded the strategy in two things: the market shape, and a first real signal from the field rather than assumptions. A short survey went to working contractors through the founding team’s network. One completed response isn’t proof — it’s a directional signal — but it lined up tightly with the problem and, more usefully, corrected one of my assumptions.
Struggle to find consistent projects?
Always
Confirms the core supply-side pain
Hardest part of your process?
Discovery — not paperwork — is the pain
Lost work because clients couldn’t see past projects?
Yes
Validates the “no portable portfolio” gap
Which feature would help most?
Portfolio showcase
Portfolio should lead the product
Time to prepare a quote?
Under 30 min
A correction — quoting is not the bottleneck
THE REFRAME
Word-of-mouth isn’t a discovery habit a better search box can beat — it’s a risk-management habit. When someone refers a contractor, they’re personally vouching and absorbing the social risk if it goes wrong. A directory of stars from strangers doesn’t replace that feeling; it’s a phone book with extra steps.
That single reframe changed the target. The product doesn’t need to out-search a referral. It needs to reproduce the two things a referral silently provides — proof and accountability — for two people who have never met.
Product strategy — key decisions
Six decisions define the product. Each is a bet with a tradeoff, not a feature wish-list.
01
Lead with trust infrastructure, not discovery
Directories already solved discovery. The defensible layer is trust: identity and certification checks, portfolios tied to confirmed clients, milestone escrow, and a reputation score weighted by completed, paid projects rather than gameable stars. Tradeoff — slower onboarding, accepted: friction here is the moat
02
Digitize word-of-mouth instead of competing with it
Early growth comes through people already inside the founding team’s real trust network, not cold ad traffic. Where possible the app shows why a professional is recommended — mirroring how early Airbnb grew: friends-of-friends before strangers.
03
Make the verified portfolio the hero of v1
Driven directly by the field signal. Photos only count once the client on that project confirms them in-app — turning a claim into proof. Voice-to-text quoting was demoted to a later accessibility aid.
04
Treat escrow as structural trust, not a payments feature
Milestone-based escrow answers the single biggest fear behind sticking with referrals: what if I pay and they disappear or do bad work. It’s the platform’s version of a referral’s unspoken guarantee.
05
Go deep in one district before going wide
Word-of-mouth is dense and local. Spreading thin across Tamil Nadu on day one dilutes the one real advantage. Winning a single district completely is when word-of-mouth starts working for the platform.
06
Validate with no code before building anything
Phase 0 runs 10–20 real client projects over WhatsApp and a spreadsheet, using the family trade network as the first verified professionals — proving demand cheaply before a line of app code is written.
The product
A two-sided mobile app — one side for clients who need work done, one for the professionals who do it — held together by a shared verification and reputation layer.
Client journey
01 Post a project — with photos, budget range and timeline.
02 Get matched with 3–5 verified professionals; browse confirmed past work.
03 Compare quotes side by side, chat directly, agree milestones in-app.
04 Pay as each milestone completes; review at close — part of a permanent record.
Professional journey
01 Build one verified profile — ID, trade proof, portfolio — that lasts for life.
02 Receive leads matched to trade and location; quote on the ones that fit.
03 Manage multiple jobs and get paid per completed milestone.
04 Grow a reputation score that compounds over time.
Feature roadmap
v1 / MVP — Verified profiles, project posting, trade + location search, in-app chat, structured reviews, WhatsApp notifications.
v2 / Growth — Milestone-based escrow payments, structured bidding / quoting, in-app project timeline tracker.
v3 / Platform — Materials marketplace, financing / BNPL for materials, insurance partnerships, AI-assisted cost estimation, tool-rental module.
Where it sits in the market
Every existing player leaves the same gap open — none is built for the multi-week, multi-trade project.
On-demand home services — Fast, funded, trusted for quick single-visit jobs. Gap: not built for multi-week projects; thin in Tier-2/3 towns.
Managed renovation platforms — Strong design + project management for interiors. Gap: premium, city-limited; skips core civil / structural work.
Directories — Wide listings, cheap to appear on. Gap: no verification, comparison, or payment structure.
Word of mouth — High trust when it works. Gap: tiny reach, unfair to newcomers, no record or recourse.
Validation approach
The strategy deliberately front-loads evidence over spend. Before any build:
• Run Phase 0 manually for 4–8 weeks — onboard the trade network, source 10–20 real client projects, track everything in a spreadsheet.
• Interview 15–20 recent renovation clients outside the family circle about how they found their contractor and what almost went wrong.
• Treat the unit economics — tiered commission — as hypotheses to test against real Phase 0 numbers, not projections to promise.
Only once Phase 0 shows repeatable demand does an MVP build — and a seed ask framed around three concrete milestones — become credible.
Reflection
What I’d carry forward
• Interrogate the incumbent behaviour, don’t just out-feature it. Reframing word-of-mouth as risk-management was the whole unlock.
• Let a single real signal correct a favourite assumption. One survey response was enough to demote a feature I was attached to.
• Design the go-to-market as part of the product. Going deep in one district is the mechanism that makes the trust layer work.
What’s still unproven
This is an early-stage concept, and I’ve kept it honest about that. Demand is validated by lived experience and one field response, not yet by a running Phase 0. Disintermediation is the hardest open question, and the answer needs real usage to confirm. The next phase is about replacing every assumption here with a real number.
@crafted by Thani
Get in touch: thanigaivasan97@gmail.com