winning-product-criteria

name	winning-product-criteria
description	Decision framework for whether a product can profitably scale on Meta Ads to Indian buyers. 7-axis scoring (each 1-5), GO threshold, and concrete tests to apply per axis. Used by product-research before any launch is approved.

Winning Product Criteria — Decision Framework

Two-layer decision:

1. Pre-screen — 13-KPI YES/NO checklist (§0). Need ≥7/13 to be eligible for depth-scoring.
2. Depth score — 7-axis 1-5 scoring (§1). Need ≥24/35 to be GO.

A product that fails the pre-screen is KILLED on the spot — don't waste research cycles depth-scoring it.

---

0. Pre-screen — 13-KPI checklist (run first)

Every candidate must hit ≥7 of 13 before it earns the depth scoring below. Score honestly — no half-points.

#	KPI	YES test
1	Small in size	Fits in a shoebox; dimensional weight under the 500g courier slab
2	Easy to ship	Not fragile · not battery-restricted · not >100ml liquid · not on Meta restricted list
3	High margin	Sell price ≥ 3.5× landed cost (will be re-verified by unit-economics hard gate)
4	Proof of concept from past winners	At least one competitor ad has been running 30+ days in Meta Ad Library, OR a near-identical SKU has scaled on Shopify/TikTok in last 12 months
5	Improves confidence	Buyer feels better about themselves after using — body, skin, hair, posture, style
6	Improves convenience	Removes a daily friction (one-handed, hands-free, faster setup, portable)
7	Saves people time	Cuts a 10-min task to 2 min, or eliminates a repeat chore
8	Solves a real problem / fills a gap	Pain stateable in one sentence: "X is annoying because Y"
9	Saves people money	Replaces a recurring spend (salon, repair, electricity, refills, subscriptions)
10	Extremely unique	Mechanism or form factor not in Amazon.in / Flipkart top 10 results
11	Improves quality of life	Sleep, comfort, peace-of-mind, ambient delight — the "luxury for cheap" feeling
12	High perceived value	Looks like ₹2,000+; priceable comfortably in the ₹699-₹1,499 sweet spot
13	Woman-dominated audience	Female buyer dominant (kitchen, beauty, kids, home decor, jewelry, fashion). India D2C buying power skews heavily female — this is the single highest-converting demographic on Meta India

Pre-screen thresholds

Hits	Verdict
10-13	Rare — jump the queue, fast-track to depth-scoring + brief
7-9	Standard candidate — proceed to depth-scoring
4-6	Marginal — attempt one angle/positioning pivot, then re-screen once
≤3	KILL — don't burn research cycles

Pre-screen rules

• KPI #4 (proof of concept) is the highest-weight signal — a YES here is worth the operator's confidence on three fence-sit KPIs. Always link the competitor Meta Ad Library URL into the Product Arsenal fb_ad_url field when this KPI is YES.
• KPI #13 (woman-dominated audience) is India-specific weight — 70%+ of impulse D2C revenue in our category mix comes from female buyers. Missing #13 requires extra-strong margin (≥4×) to compensate.
• KPIs #1, #2, #3 are structural — a NO on any of these makes the product nearly un-launchable regardless of the other 10. Treat a NO on a structural KPI as a soft kill unless margin is exceptional.
• Pre-screen is a fast, ~2-minute pass. If you're agonizing over a YES/NO, mark it NO and move on.

---

1. Depth scoring — 7-axis (run only after pre-screen ≥7/13)

Score each candidate 1-5 on every axis. Total ≥24/35 → GO. Below that, HOLD or KILL.

Axis	5 (excellent)	1 (bad)
Margin	≥3.5x landed cost incl. ads	<2x; thin sliver after returns
Wow factor	Demoable in 5 sec, "I need this" reaction	Boring, commodity-feeling
Broad appeal	30M+ Indian buyers in target	Narrow niche <2M
Video-friendly	Visual demo carries the sell	Needs explanation, no demo
Light + small	<500g, <30cm, courier-safe	Heavy/fragile/large/restricted
Not big-box	Not on Amazon top 3 / Flipkart top 3	Saturated and undercut
India COD-fit	Sweet spot ₹699–₹1,499 + low RTO category	>₹2,500 for cold COD audience

Per-axis tests

Margin (target ≥3.5x)

• Landed cost = product + supplier shipping + packaging + courier-to-customer + RTO loss provision
• RTO provision = (landed cost) × (expected RTO %) (default 25%)
• Required margin to scale = 3.5x landed cost = sell price ≥ 3.5 × landed
• If supplier pricing is negotiable, push for 15-20% off list before scoring

Wow factor

• 3-second test: show an unedited product photo to a non-buyer. If reaction isn't "what is that / I want one", score ≤2.
• Higher scores for: solves visible pain, novel mechanism, satisfying-to-watch in motion, transformation (before/after)

Broad appeal

• Estimate via Meta Audience size for relevant interests (use ads_insights_advertiser_context)
• 30M+ → 5; 10-30M → 4; 5-10M → 3; 2-5M → 2; <2M → 1
• Niche products can win but need premium pricing to offset smaller pool

Video-friendly

• Can you demo it in a 5-second clip? Yes → 4-5
• Does the demo "satisfy" (oddly satisfying / before-after / transformation)? → 5
• If you need text overlays to explain → 2-3
• If only static product shots work → 1-2

Light + small

• ≤500g + dimensional weight under courier slab → 5
• 500g–1kg → 3
• 1-2kg → 2

• 2kg or fragile (glass, ceramics) or restricted (battery, liquid >100ml) → 1

• Heavier = higher courier cost = thinner margin + higher damage rate

Not big-box

• Search exact product on Amazon.in and Flipkart
• Not in top 10 results → 5
• In top 10 but priced >2x our sell price → 4
• In top 10 with similar price → 2
• Top 3 Prime + cheaper → 1 (kill)
• Indian buyers will Google before COD; if they find your product cheaper on Amazon, they cancel

India COD-fit

• ₹699-1,499 + low-RTO category (kitchen, gadgets, accessories) → 5
• ₹1,500-2,499 + medium RTO → 3-4
• ₹2,500-4,999 → 2 (only with prepaid push)

• ₹5,000 cold COD → 1

Hard kills (instant disqualify regardless of score)

• Health/medical claims (diabetes, hair regrowth, fat burning, libido, etc.)
• Branded/trademarked products (Apple-style chargers, fake brands)
• Restricted categories on Meta (weapons, supplements, surveillance)
• Items that fail Indian customs or BIS certification (some electronics)
• Products requiring regulatory licenses you don't hold (cosmetics → CDSCO, food → FSSAI)
• Items with known IP infringement (popular character merchandise without license)

GO / HOLD / KILL

A candidate must clear both layers:

• GO Pre-screen ≥7/13 and depth score ≥24/35 and no hard kill and unit-economics hard gate passes (NET PROFIT per total order > 0 at FAD 0.70 + CPP = 8% of SP) → push to launch brief
• HOLD Pre-screen ≥7/13 but depth 18-23, OR pre-screen 4-6 → identify the weakest signals, attempt one lift (negotiate margin, alt supplier, repackage angle, sharper hook); re-screen + re-score
• KILL Pre-screen ≤3, OR depth <18, OR any hard kill, OR unit-economics gate fails → drop entirely, log in Notion "Killed" view with reason (which layer failed)

Notes for product-research agent

• Always run pre-screen FIRST. Report the 13-KPI hit-count and which KPIs were YES vs NO before doing anything else.
• If pre-screen fails (<7/13), STOP — don't depth-score, don't write a brief. Log KILL + reason and return.
• If pre-screen passes, then do depth-scoring and the unit-economics gate.
• Always show the score breakdown, not just the total.
• Always state the two weakest signals (one from pre-screen NOs, one from lowest depth axis) and the operator action that would raise them.
• If GO: produce the launch brief. If HOLD: produce the lift plan. If KILL: log and move on.