What a performance marketing agency costs an Indian D2C food brand

Every founder asking this question wants one number. There isn’t one, and any agency that gives you a headline price without asking about your SKU count, channel mix and creative volume is quoting a package rather than a scope.

What we can do is show you what is publicly published, by whom and when, explain what actually moves the number, and give you a way to test whether the proposal on your desk is sane for a food brand at your stage.

One disclosure first. We do not publish The Shizz’s rates, and this article does not sneak them in. The reasons are at the end. What we do publish is that the Growth Audit is free.

What the published ranges actually say

Three Indian agencies have published dated pricing guides in 2026. Their numbers are attributed below exactly as each source states them. Read each as one firm’s view of the market, not as market truth.

upGrowth, published 20 April 2026

In Performance Marketing Retainer Pricing India (2026), author Amol Ghemud sets out three structures: a flat retainer of ₹1.5L to ₹6L+ a month, a percentage of ad spend of typically 10 to 15 percent, or a hybrid of a flat floor of ₹1.5L to ₹3L plus a percentage, whichever is higher.

The article’s benchmark by ad-spend tier, as published:

• Ad spend under ₹5L a month: a quality retainer of ₹1.2L to ₹2L flat. The article states that above ₹2L at this spend level, “the math doesn’t work”.
• ₹5L to ₹15L a month: ₹2L to ₹3L flat, or ₹1.5L plus 10 percent.
• ₹15L to ₹50L a month: ₹3L to ₹5L flat, or ₹2L plus 12 percent.
• ₹50L to ₹2 Cr a month: ₹5L to ₹10L flat, or ₹3L plus 10 percent.
• Above ₹2 Cr a month: typically ₹10L to ₹20L flat, or ₹5L plus 8 percent.

A separate table in the same article prices by tier rather than spend: ₹75k to ₹1.5L a month for media spend up to ₹10L, ₹1.5L to ₹4L for ₹10L to ₹50L, and ₹4L to ₹10L+ above ₹50L. The article also states a rule worth remembering: never agree to a pure percentage structure below ₹20L a month of ad spend, because the percentage underfunds the team.

PeakPilots, published 5 June 2026

In D2C Marketing Agency Pricing in India: What to Expect in 2026, author Dhaval Vaghasiya publishes a noticeably lower set of ranges, scoped specifically to D2C:

• Early scale, ₹1L to ₹3L monthly ad spend: a typical retainer of ₹40k to ₹80k.
• Growth, ₹3L to ₹8L monthly ad spend: ₹80k to ₹1.5L.
• Scale, ₹8L+ monthly ad spend: ₹1.5L+, scoped.

The same article states that most serious D2C agencies in India require ₹1L+ a month of active ad spend on Meta or Google, and describes those figures as directional, to be replaced by a scoped proposal.

Adservex, published 13 June 2026

In Digital Marketing Agency Pricing in India: 2026 Guide, the Adservex team prices by agency tier rather than by client spend:

• Boutique and solo specialists: ₹35,000 to ₹1,00,000 a month.
• Mid-tier independent agencies: ₹1,50,000 to ₹4,00,000 a month.
• Premium performance agencies in Mumbai, Bengaluru and Delhi NCR: ₹4,00,000 to ₹10,00,000+ a month.
• Network agencies: ₹8,00,000 to ₹25,00,000+ a month.

It puts single-channel paid management at ₹75,000 to ₹2,50,000 a month plus a media-spend fee, full-funnel paid at ₹2,50,000 to ₹8,00,000 plus a fee, and the media-spend fee itself at 8 to 15 percent, sliding to 8 to 10 percent above ₹50L a month of spend. It also states that Mumbai, Delhi NCR and Bengaluru run 20 to 35 percent higher than Pune, Hyderabad, Chennai, Ahmedabad or Kolkata for the same scope.

They disagree, and that is the useful part

At ₹3L of monthly ad spend, PeakPilots points at roughly ₹80k of retainer and upGrowth points at ₹1.2L to ₹2L. That is not a contradiction so much as a difference in what the word “agency” is being made to mean: a lean D2C team running two channels is a different product from a full-funnel engagement with a named senior strategist, an in-house creative pod and attribution engineering. Both can be honest. Neither is a benchmark you can hold a vendor to.

Use them the way you would use three property listings on the same street: to establish a plausible band, not a price.

What actually drives the number

Scope, not seniority theatre. In roughly the order they add cost:

• Creative volume. This is the largest line in most performance retainers, and the upGrowth article says so explicitly, describing 2026 Meta and YouTube performance as a creative arms race and pricing a mid-tier retainer around roughly 30 variants a month. Whether that production is in-house or outsourced changes both the cost and the turnaround.
• Channel count. Meta only is one thing. Meta plus Google plus Amazon plus Blinkit and Zepto is four ad consoles, four reporting logics and four sets of creative specs. Food brands accumulate channels faster than most categories.
• Spend volume. Bigger budgets need more senior hands, more frequent review and more structural work, though the relationship is nothing like linear.
• Tracking and attribution. Server-side events, Conversions API, enhanced conversions, offline imports. This is engineering work, and it is the line item most often quietly dropped from a cheap proposal.
• Everything around the ad account. CRO, retention flows, marketplace listings, landing pages. Optional in the contract, rarely optional in reality.
• City. Adservex puts the metro premium at 20 to 35 percent for identical scope.

The four pricing models, and what each one rewards

Flat retainer

A fixed monthly fee for a defined scope, with media paid separately by you. Predictable, easy to budget, and the agency’s income does not rise when your spend does. The failure mode runs both ways: scope creep eats the agency’s margin until service quality drops, or a vague statement of work quietly shrinks as the agency signs new clients. Fix it by writing monthly minimums into the contract: number of creatives shipped, tests run, reports delivered, named people and hours.

Percentage of ad spend

Usually 10 to 15 percent in India per upGrowth, or 8 to 15 percent as a media fee on top of a retainer per Adservex. Be clear-eyed about this one: it pays the agency more when you spend more, and it pays them nothing extra for making the same spend work harder. That is a real incentive problem, not a theoretical one, and it is worst in food, where thin contribution per order means the difference between scaling and over-spending is a narrow band.

It also fails at the bottom. Twelve percent of ₹8L is ₹96,000, which will not fund a senior strategist and a creative pipeline. That is the reasoning behind upGrowth’s rule against pure percentage below ₹20L of monthly spend. If you do use this model, cap it, tier it so the rate falls as spend rises, or pair it with an efficiency term.

Hybrid: floor plus percentage

A flat floor that funds the team, plus a percentage that scales with growth, with the client paying whichever is higher. upGrowth describes this as its default above ₹15L of monthly spend and gives a worked example of a ₹2L floor plus 12 percent. It fixes the underfunding problem at low spend and the volatility problem at high spend, but it inherits the incentive to grow spend, so the same caps and tiers apply.

Performance-linked and revenue share

Adservex describes revenue share as rare, at 3 to 8 percent of incremental revenue, and upGrowth notes a performance-only variant taking 10 to 20 percent of net revenue generated. Both sources warn about the same thing, and they are right to: an agency paid only on a metric will optimise that metric. In practice that means discount-led offers that hit revenue and destroy margin, brand-term search hoovering up conversions that were coming anyway, and attribution arguments in month three. If you want alignment, a modest bonus on top of a real base is the version that works.

How to judge the quote in front of you

Four steps. None of them require knowing what anyone else pays.

1. Convert every quote into two ratios

First, fee as a percentage of planned ad spend. That makes a flat retainer and a percentage model directly comparable. Second, and more important, fee as a percentage of the contribution margin that spend is expected to produce. upGrowth’s published rule of thumb is that a retainer sitting at 3 to 5 percent of contribution margin is healthy, and above 10 percent means either the fee is too high or the performance is too low. Adservex frames the same test differently: total agency cost of 8 to 18 percent of the revenue or pipeline it is expected to influence.

2. Do the food-brand version of that arithmetic

Take an illustrative brand, not a benchmark: average order value ₹700, gross margin 45 percent, and roughly ₹120 a shipment in shipping, payment and COD handling. That is about ₹195 of contribution per order before media. At ₹5L of monthly ad spend and a 3× ROAS you are doing ₹15L of revenue, about 2,140 orders, and roughly ₹4.2L of contribution. Now put the proposed fee against that number. A fee that eats a third of your contribution has to produce a step change to be worth it, and both published sources would call that ratio a problem. Run the same sum with your own AOV and margin before the second meeting.

3. Divide the fee by the output

How many new creative concepts a month, not variations? How many channels, actively managed rather than listed? Who is on the account, named, and for how many hours a week? A fee is only expensive relative to what it buys, and this is where most quotes fall apart.

4. Read what is excluded

The usual gaps: creative production billed separately, tracking and Conversions API setup as a paid project, landing pages out of scope, marketplace and quick commerce out of scope, and reporting limited to platform dashboards. None of these are wrong to charge for. All of them are wrong to discover in month two.

Red flags that hold across all three sources

• Full-funnel growth quoted under ₹50,000 a month. Adservex calls this freelancer scope dressed as agency scope.
• Twelve-month lock-ins on a first engagement, when a 90-day pilot would do.
• Percentage of spend with no cap, no tier and no efficiency term.
• No mention of tracking setup anywhere in the proposal.
• Creative quoted as two to four ads a month.
• Reporting that stops at platform ROAS, with no contribution or CAC view.

Why we do not publish our own rates

Two honest reasons. A single published number would be wrong for almost everyone reading it: a four-SKU ghee brand on Shopify and a forty-SKU snacking brand across Meta, Amazon and three quick-commerce platforms are not the same job, and a headline price would either flatter us or scare off a brand we could have helped. And a rate without a scope is not information, it is anchoring.

What we publish instead is the work. Kroslo doubled ROAS in 25 days on the same budget by removing waste rather than adding spend. Aazol cut acquisition cost by nearly 70 percent and moved past ₹22.5L a month over four months. Across six years the aggregate is 160+ brands, ₹150 Cr+ of ad spend managed, ₹450 Cr+ of revenue attributed and a 3.8× average ROAS. The Growth Audit is free, and you get the roadmap whether or not you hire us.

A fee is not expensive or cheap on its own. It is expensive or cheap relative to the contribution margin it moves, and that is a number only you can put in front of it.