Saral Shiksha Yojna
Courses/Technology Product Entrepreneurship

Technology Product Entrepreneurship

CS9.424
Ramesh Loganathan + Prakash YallaMonsoon 2025-264 credits

Defining the Problem, Quantifying Implications, Bias-Free Validation

NotesStoryCaveman
Unit 2 — Problem Analysis

Ugg no marry own spear

Long ago Ugg carve beautiful spear. Ugg love spear. Then wise Priya-woman ask: "Ugg — what beast you even hunting?" Ugg go quiet. Ugg fall in love with the spear, never ask why tribe hungry.

Levine's Law — rule from Uri Levine, chief who built the Waze path-finder: love the problem, not the solution. Pretty spear for pretend beast still starve the tribe. Ugly-but-working trap for real beast feed them, then get better. Hunt the hurt, not the tool.

Levine's Tree — see a tree. Roots in dirt = the Problem (why tribe hungry). Trunk = the Solution (the plan: dig pit, set trap). Leaves up top = the Product (thing the tribe holds and sees). Weak root, whole tree fall — no pretty leaf save it. Draw fast: roots, trunk, leaves, bottom to top. Ugg warn: Ugg spent two weeks decorating leaves and never check the root!

Three ways to grunt the same hurt

Climb the ladder:

Level 1 — Vague grunt. "People need food." Say nothing. Who? Why? How bad? Baby-founder stop here.

Level 2 — One-cave grunt. "I busy hunter, I no time gather berry, this make me sad." Better — got the feeling. But one cave only, not whole valley.

Level 3 — the good grunt (4Ws + When). "Our busy hunter-woman cannot get healthy food on long-hunt days because she gone dawn-to-dark. Tribe need quick way to hand her good food." That the shape the chief-teacher want.

Means: name WHO (which tribe-person), name WHEN/WHERE (the long-hunt day, the cave), name what they LACK (the hole, the hurt). All three, or it not Level 3.

Five questions before any hunt — and the keystone

5Ws Framework — five stone questions round the fire:

  • Who — which tribe-people feel the hurt? Not just the eater; everyone in the circle.
  • What — how things ARE versus how things SHOULD be. The gap between = the unmet need. Remember that grunt.
  • When — how often, and in what season? Once-a-winter hurt is a different beast from every-sun hurt.
  • Where — which cave, which valley? A Tier-1 cave is not a far-mountain cave.
  • Why — the Critical Gap. Why this hurt worth the hunt? How BIG?

Why is the keystone — the one stone that hold the whole arch up. Every trade-chief ask "so what?" at the end. No good Why, the other four W no save you. Ugg warn: no spread grunts even across all five — make Why the LONGEST, and put a number on it. "₹1.5 lakh crore of wasted fire-power a year" beats "people annoyed."

The tribe is not one person

Big trap: Stakeholder Ecosystem — the eater is not the buyer, and the buyer is not the yes-sayer. Seven roles orbit any big-cave trade:

Decision Maker (chief, final yes or kill), Influencer (whisper-elder, sway without power), Buyer (carry the shiny stones, seal the trade), Manager (watch over the users), Operator / End User (actually throw the spear, day to day), Evaluator (sniff-test it before the trade), CFO (guard the shiny-stone pile).

Salesforce cave: spear-thrower = User, hunt-VP = Buyer, shiny-stone-guard = Decision Maker. Ugg warn: if your grunt name only the tired spear-thrower, you hunt the WRONG cave! Each role feel a different flavor of the same hurt — so write a separate hurt-grunt for all seven.

How big the hurt, really — the three-question funnel

Naming the hurt is half. Proving it big enough to feed a tribe is the other half. Implications Triad — three questions, say each out loud:

Magnitude — how big per bite? What happen if no fix, counted in meat, not moaning. Fire-tale: China bad-air = dead people + 20,035 gigagram of lost crop = 267 billion Yuan of hurt. Say the number, not "big problem."

Frequency — how often the hurt come? Every-sun hurt = tribe pay again and again. Once-a-winter = one trade only. High-often = sticky = food keep coming.

Population — how many caves feel it? A big hurt for only fifty caves is a small hunt. Ugg warn: no mix up Frequency and Population! Frequency = how often per cave; Population = how many caves.

Stack them:

That the whole size of the hurt-mountain — call it TAP. Ugg warn: no double-count! If one cave hurt many times a year, that is frequency — no also count each hurt as a new cave.

Medicine-root or sweet berry?

A Painkiller is the medicine-root the tribe grab NOW to stop real pain — they pay quick. A Vitamin is a sweet berry, nice-to-have, maybe grab if they remember.

Means: small hurt OR rare hurt, and it just a berry. Trade-chiefs feed painkillers, not berries. Scratch the 2×2 in the dirt: big-hurt + often = painkiller (top corner, sweet spot); big-hurt + rare = one-time trade (mammoth-insurance); small-hurt + often = tolerated itch (slow river); small + rare = berry. Only the top corner feed a growing tribe. Ugg warn: no call your berry a medicine-root just because YOU want it to be! Be honest on Magnitude and Frequency — reframe to find the painkiller side, or kill the idea.

Ask other caves — but no crooked questions

Every number so far is just Ugg's guess. Go ask real caves — but no poison the well.

Survey design — three fire-rules:

1. No leading question — "you like Ugg idea?" already bake the answer in. Crooked. 2. Ask past behaviour — the Past-Behaviour Question. Not "would you hunt here?" but "when did you LAST hunt here?" Real footprints beat pretend ones. Tribe lie about the future, tell truth about the past. 3. Quantify the pain — not "is this bad?" but "how many suns did you lose?" Numbers force honest.

Means: keep past-tracks and counted-pain; throw out crooked leading questions. Say it once round the fire: Bias is the enemy of validation. The Bias-Trap drill: ❌ "Would you pay for store-manager magic?" versus ✅ "How many hours a week your store-managers now spend on hand-counting?" Factual, counted, no push.

Validation Workflow: make survey → pick target (your Who from the 5Ws) → collect ≥ 20 answers → chew results, fix your grunt. Ugg warn: even good past-questions over-count — real buying is half or less of what caves SAY, so shrink the signal. And never pitch your spear in these talks — this is footprint-tracking on the hurt; spear-selling come later.

Your own opinion is nothing. Only the cave's real tracks matter.

Ugg remember

  • Love the problem, not the spear. Levine's Tree: root = Problem, trunk = Solution, leaf = Product. Weak root, tree fall.
  • Climb to Level 3: Who + Context + Deficit. Then the 5Ws — and make Why the longest, for it is the keystone that answers "so what?"
  • The tribe is seven roles, not one: Decision Maker, Influencer, Buyer, Manager, Operator, Evaluator, CFO — write a hurt-grunt for each.
  • Painkiller = High Magnitude × High Frequency. Size the whole hurt = Magnitude × Frequency × Population (TAP). Berry no get fed.
  • No crooked questions: past behaviour + counted pain − leading. Collect ≥ 20 answers. Bias is the enemy of validation.
End of storyUnit 2 — Problem Analysis · Defining the Problem, Quantifying Implications, Bias-Free Validation