The shape of the model
The calculator starts from the revenue you want, converts it into the number of units ads must actually drive, checks whether your unit economics can afford that, prices those units against live India auction benchmarks, inflates them for the frictions specific to your brand, then spreads the result across a realistic ramp.
Revenue target → units ads must drive (Layer 1) → maximum spend you can afford (Layer 2) → cost to buy those units (Layer 3) → friction adjustment (Layer 4) → month-by-month ramp (Layer 5) → three scenarios (Layer 6)
Layer 1 · Demand model
Not all revenue is paid revenue. A share of sales arrives organically — through search rank, repeat buyers and brand recall. The model estimates the paid share as an ad dependency figure driven by brand stage, then softened by average selling price: higher-priced products convert on fewer, more considered clicks.
| Brand stage | Ad dependency | Reasoning |
|---|---|---|
| New launch | 70% | No rank, no reviews, no repeat base |
| Growing | 45% | Some rank and returning buyers |
| Established | 25% | Organic and repeat carry most volume |
| Average selling price | Price multiplier |
|---|---|
| Under ₹300 | 1.10 |
| ₹300–₹799 | 1.00 |
| ₹800–₹1,499 | 0.95 |
| ₹1,500–₹2,999 | 0.90 |
| ₹3,000+ | 0.80 |
Returns are then added back. A sale that comes back is a unit ads paid for twice, so gross units are grossed up by an effective return rate — the category baseline scaled by your COD share, because cash-on-delivery orders are returned far more often.
gross units = (revenue × ad dependency × price multiplier) ÷ ASP effective return = category return rate × (0.5 + COD share) units to buy = gross units ÷ (1 − effective return)
Category return baselines: Fashion 22%, Home 12%, Electronics 12%, Beauty 10%, Health 10%, Food 6%. Capped at 60%.
Layer 2 · Break-even guardrail
Before pricing a single click, the model asks what you can afford. Break-even ROAS is the inverse of gross margin; the target ROAS adds a 20% contribution buffer on top so ads are not merely washing their own face.
break-even ROAS = 1 ÷ gross margin target ROAS = break-even ROAS ÷ 0.80 max spend = revenue target ÷ target ROAS
If the modelled budget exceeds that ceiling, the calculator says so explicitly. That is not a rounding problem — it means the plan requires spending more than the margin supports, and price, margin or target has to move first.
Layer 3 · Platform benchmarks
Units are split evenly across the performance platforms you select, then priced with India category benchmarks. Marketplace and search platforms are modelled on cost per click and conversion rate; Meta on cost per mille, click-through rate and conversion rate.
CPC platforms spend = units × (1 ÷ CVR) × CPC
Meta impressions = units ÷ CVR ÷ CTR
spend = (impressions ÷ 1000) × CPM| Platform | Conversion rate band |
|---|---|
| Amazon | 8–12% |
| Flipkart | 6–10% |
| 2–4% | |
| Meta | 1.5–2.5% (CTR 1.0–1.8%) |
Quick commerce does not run on an auction you can budget into gradually. Blinkit, Zepto and Instamart are modelled as flat monthly commitments and shown separately, never blended into the CPC maths.
| Platform | Monthly commitment | Structure |
|---|---|---|
| Blinkit | ₹2.0L–₹3.0L | Flat monthly, auction-based Seller Hub inventory |
| Zepto | ₹1.5L–₹2.5L | Flat monthly, or bundled in a ₹5–6L onboarding package |
| Instamart | ₹2.66L–₹3.33L | ₹8–10L quarterly, shown as monthly equivalent |
Layer 4 · Friction multipliers
Benchmarks describe an average advertiser. You are not average. Four multipliers compound onto every performance platform figure.
| Driver | Range | What it represents |
|---|---|---|
| Brand stage | 1.00 – 1.35 | Conversion-history premium the auction charges new accounts |
| Listing quality | 0.90 – 1.55 | Weak content wastes paid clicks that never convert |
| Seasonality | 0.85 – 1.45 | Festive auction inflation vs the January–February trough |
| Category competition | 0.95 – 1.20 | How crowded the auction is in your category |
The listing-quality multiplier is the one you control fastest. Moving from Average to Good removes 25% of required spend with no change to the media plan at all — which is why the calculator surfaces it as an improvement lever rather than burying it.
| Listing score | Rating | Multiplier |
|---|---|---|
| 85–100 | Excellent | 0.90 |
| 65–84 | Good | 1.00 |
| 45–64 | Average | 1.25 |
| Below 45 | Poor | 1.55 |
| Months | Season | Multiplier |
|---|---|---|
| October–November | Festive peak | 1.45 |
| September, December | Festive shoulder | 1.25 |
| March–August | Normal | 1.00 |
| January–February | Cheapest auction | 0.85 |
Layer 5 · Phased ramp
Spending the full budget in month one buys expensive data, not sales. Algorithms need conversion volume before efficiency arrives, so the recommendation is delivered as a ramp with ROAS expectations that start deliberately low.
| Phase | Spend | Expected ROAS | What is happening |
|---|---|---|---|
| Months 1–2 · Learning | 55% of budget | 40% of target | Buying information; algorithms need conversion volume |
| Months 3–4 · Optimisation | 80% of budget | 70% of target | Winners identified, losers cut |
| Month 5 onward · Scale | 100% of budget | 100% of target | Organic rank now supports paid; blended efficiency improves |
Layer 6 · Three scenarios
A single number implies a precision the auction does not offer. The calculator returns three, and leads with the conservative one, because a plan that only works in the best case is not a plan.
| Scenario | Spend | ROAS | Assumption |
|---|---|---|---|
| Conservative (recommended) | +15% | 75% of target | CPC 20% above benchmark, CVR 15% below |
| Expected | Modelled budget | Target | Benchmarks hold |
| Stretch | −10% | 130% of target | Strong creative, listing score 85+, favourable auction |
All outputs are rounded to the nearest ₹1,000.
What this model does not know
It does not know your creative quality, your competitor's pricing move next week, whether you will go out of stock mid-campaign, or how your specific account has been scored historically by each platform. Those are the variables that separate a good plan from a good result.
Estimates are directional, built on India category benchmarks and adjusted for your brand stage, listing quality, category competition, and season. Actual performance depends on creative quality, price competitiveness, stock availability, and auction dynamics. Budgets are reviewed and revised monthly against live performance data.