Documentation

How the Ad Budget Adviser works

Every number the calculator returns comes out of six sequential layers. Nothing is a rule-of-thumb percentage of revenue. This page documents each layer, the benchmarks it uses, and where the model can be wrong.

← Back to the calculator

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 stageAd dependencyReasoning
New launch70%No rank, no reviews, no repeat base
Growing45%Some rank and returning buyers
Established25%Organic and repeat carry most volume
Average selling pricePrice multiplier
Under ₹3001.10
₹300–₹7991.00
₹800–₹1,4990.95
₹1,500–₹2,9990.90
₹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
PlatformConversion rate band
Amazon8–12%
Flipkart6–10%
Google2–4%
Meta1.5–2.5% (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.

PlatformMonthly commitmentStructure
Blinkit₹2.0L–₹3.0LFlat monthly, auction-based Seller Hub inventory
Zepto₹1.5L–₹2.5LFlat 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.

DriverRangeWhat it represents
Brand stage1.00 – 1.35Conversion-history premium the auction charges new accounts
Listing quality0.90 – 1.55Weak content wastes paid clicks that never convert
Seasonality0.85 – 1.45Festive auction inflation vs the January–February trough
Category competition0.95 – 1.20How 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 scoreRatingMultiplier
85–100Excellent0.90
65–84Good1.00
45–64Average1.25
Below 45Poor1.55
MonthsSeasonMultiplier
October–NovemberFestive peak1.45
September, DecemberFestive shoulder1.25
March–AugustNormal1.00
January–FebruaryCheapest auction0.85

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.

PhaseSpendExpected ROASWhat is happening
Months 1–2 · Learning55% of budget40% of targetBuying information; algorithms need conversion volume
Months 3–4 · Optimisation80% of budget70% of targetWinners identified, losers cut
Month 5 onward · Scale100% of budget100% of targetOrganic 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.

ScenarioSpendROASAssumption
Conservative (recommended)+15%75% of targetCPC 20% above benchmark, CVR 15% below
ExpectedModelled budgetTargetBenchmarks hold
Stretch−10%130% of targetStrong 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.