Building AOPs That Work: Real Lessons from Seasoned D2C Founders

In a market where quick commerce can transform demand overnight and procurement windows make or break margins, the Annual Operating Plan has evolved from a fundraising document into something far more powerful — a daily operating system for the best D2C brands.

Unlike traditional retail or FMCG, D2C AOPs must simultaneously manage channel-level P&Ls (from your own website to quick commerce), category-wise growth trajectories, and SKU-level procurement decisions — all while maintaining the flexibility to pivot when Blinkit adds 200 dark stores or Instagram changes its algorithm.

To deconstruct what a living, breathing AOP looks like, in this Field Notes, we tap the lived experience of two D2C founders, Romita Mazumdar, founder of Foxtale (₹400+ crore omnichannel beauty brand), and Abhishek Agarwal, founder of Farmley (₹400+ crore healthy snacking brand).

In the next few scrolls, you’ll find an insider’s account of building and managing AOPs, including:

• How founders move from gut to frameworks as they scale.
• Scenario planning, structure, and why every row in your plan needs a true owner in your org.
• Building channel, category, and SKU-level forecasts across D2C, quick commerce, offline, and B2B
• What real forecasting, controls, and course correction look like week-to-week
• Practical lessons on working capital moats, managing marketing budgets, and ensuring buffers in your planning.

Let’s begin with how best founders evolve their mindset.

From Startup Instinct to Scale-Ready Systems

Romita recalls their first year was about surviving and hustling, not building grand plans. The AOP was a fundraising document, but daily work was driven by bottom-up planning and instinct.

“The fiscal year 23 AOP process was bottoms up, starting monthly. Every 20^th of the month, I would plan what the next month would look like. I’d analyze how many customers visited my website and engaged on other channels. My business plan was driven by instinct — a month-to-month exercise that taught us to use today’s demand to predict tomorrow’s.”

As Foxtale scaled, discipline became non-negotiable. ​“When we were a 20-person team, I’d work with them till 1:00 am and then grab a beer. I can’t do that anymore — I don’t know every employee. The same applies to business plans: you move from being a founder who says, ​‘This is my dream,’ to building a plan that reflects how that vision is actually shaping up.”

Abhishek’s journey to structured planning took a different route. For the first three years, Farmley was purely a B2B company: ​“We were building the supply chain, infrastructure, and understanding food infra.”

The B2C shift made planning critical. ​“You have to plan for Diwali in November. To support the infrastructure we built, we had to build season-to-season AOPs — thinking three years ahead, not three weeks.”

Both founders hit the same inflection point: when ambition — ​‘I want to do ₹250 crore’ — must reconcile with reality — ​‘My monthly demand says I can do ₹150 crore.’ That’s when AOP becomes crucial to success.”

Build Your Mental Model Before Your Spreadsheet

Before building their AOPs, both founders established clear mental models — Abhishek with his three-year profitability anchor, Romita with her daily-driver philosophy.

Start At the End: Abhishek’s Three-Year Framework

Abhishek’s framework starts with a counterintuitive move: planning three years out to get year one right. ​“At Farmley we have a 3‑year AOP planning framework, not just for the coming year. This 3‑year framework has significantly improved our 1‑year planning.”

This requires clarity on direction: ​“The objectives should be clear for three years — revenue, product range, profitability metrics. Where do you want to reach in the next three years?”

He starts each AOP by anchoring on profitability: ​“The final profitability matrix is the key. If today you’re a minus 5% EBITDA company or 0% EBITDA, where do you want to go? What level of profitability do you want to hit?”

From this anchor, Abhishek builds scenarios — but with a crucial insight: don’t get stuck with traditional definitions of optimistic vs. pessimistic. ​“What’s optimistic and pessimistic depends on your objective. If my objective is to get profitable, then 20% growth with EBITDA profitability can be optimistic. The pessimistic case? Growing only 15% and missing profitability targets. So what’s optimistic or pessimistic depends on the year’s objective.”

To build these scenarios, he suggests three questions in order:

• How much do I want to grow?
• How profitable do I want to be?
• What do I want to change in the org?

Once answered, the real work begins: making concrete trade-offs. ​“You get combinations of growth scenarios with profitability scenarios. You can be a 15% growth company at 0% EBITDA; a 30% growth company at minus 5% EBITDA; a 5% growth company, but EBITDA profitable. The AOP exercise forces you to work through these trade-offs.”

Romita’s Daily-Driver Philosophy

Romita’s approach starts with a conviction: ​“AOP is not a financial exercise. It’s a business-led, business-driven exercise.” Unlike founders who treat AOPs as board documents, she sees it differently: ​“For me it’s a daily exercise. I need to know exactly where sales are coming from, which product, which consumer tier, what RoAS each is transacting at, where demand generation is happening.”

To maintain this detail at scale, ownership must be distributed. ​“AOP is not my document alone. Every channel head owns their slice, from gross revenue to RoAS. We track everything — which influencer, which region, which category drives outcomes.”

Yet granular tracking needs something more: ​“The science has to be there, but there must be a marriage between ambition and demand.”

She’s learned this balance is non-negotiable: ​“You can’t bluff the organization or your investors. You need real numbers and narrative both, and you cannot do it without a solid AOP exercise.” For Romita, the AOP isn’t about choosing between founder instinct and market data — it’s about harnessing both to build something bigger.

The Nuts and Bolts of an Effective AOP

Both founders build their AOPs differently — Romita through channel-by-channel P&Ls, Abhishek through category-first planning — but both create the same thing: complete visibility into every revenue and cost driver.

Romita’s AOP Construction: Step-by-Step

1. Channel-Wise, Bottom-Up Planning

Each channel (D2C, marketplaces, QCom, offline, B2B) builds its plan bottoms-up through dedicated P&L heads who own everything from gross revenue to CM3.

“We decide channel-wise. Every channel does its bottom-up approach to achieve gross revenue, which is GMV minus discounts. They calculate units sold and ASP, then discount is backward calculated and guardrailed.”

The ownership structure is clear: ​“Website, B2B marketplaces, B2C marketplaces, quick commerce, and offline, each has a P&L head. Channel head has to come up with an equation. Once I know it, I review what to increase or decrease.”

2. Balance the Revenue Equation

Once channel heads build their initial plans, Romita enters into detailed discussions and reviews where discount rates, marketing budgets, and RoAS targets are all variables.

“Every channel head takes full marketing responsibility and tells me: this is the spend needed to reach this gross revenue metric, and this is the RoAS I’ll operate at.”

When projecting RoAS improvements, specifics are mandatory. ​“The levers must be clearly broken down. Has conversion increased? Has AOV increased? Why will your RoAS increase?” Each assumption must tie to a concrete driver.

3. Build Channel-Wise Gross Margins & Costs

After balancing the revenue equations, the next step is mapping the true cost structure of each channel. For each channel, costs like commissions, payment gateway, logistics, and warehousing are forecasted separately.

“We do channel-wise gross margins because each channel has different costs. Marketplaces have high commissions while websites have no commissions but payment gateway charges.”

This granularity is critical for D2C operations. ​“For logistics costs, we analyze website orders regionally. The channel breakdown shows product sales and order origins.” This matters because a ₹500 order from Mumbai might be profitable while the same order from Northeast India might not — warehouse locations and shipping zones change the unit economics entirely.

4. Calculate Contribution Margins Layer by Layer

With channel costs mapped, Romita systematically builds contribution margins, creating transparency into exactly where profitability comes from.

“I know which warehouse will have what volume. Each warehouse has a different operating cost, so I calculate the weighted average warehouse cost. Every region and distance from that hub has different logistics costs — zone A, zone B pricing.”

This cost modeling produces CM1 — the first layer of contribution margin after direct costs.

The progression continues logically: ​“CM2 comes next because marketing cost has already been given by each channel head. That leads to post-marketing costs, which is CM3.”

5. Going from CM3 to EBITDA

The final bridge from CM3 to EBITDA requires planning the two largest remaining cost centers: payroll and R&D. Both are treated as controlled investments rather than variable costs.

For payroll planning: ​“Every team has a productivity target, number of people, and average salary. The targets will differ for every team.” Sales teams might have revenue per head targets, while tech teams have different metrics. Notably, ops teams are more predictable: ​“Ops team has a fixed cost because the needed personnel per warehouse square feet is clear.”

R&D follows a different logic entirely. R&D investment ties to the product pipeline: ​“R&D is driven by the number of products you plan to launch. A great scientist can only make four to five formulations a year.” Importantly, ​“R&D is a control function, not a growth function. We don’t just add people — it’s driven by the number of products to be launched.”

This disciplined approach to fixed costs is what bridges the gap from CM3 to sustainable EBITDA.

Abhishek’s Category-First Method

Once Abhishek establishes his 3‑year profitability trajectory and answers his three fundamental questions, the AOP construction follows a systematic sequence:

1. Set Category-Wise Growth Targets

Abhishek divides his portfolio into growth buckets. ​“I have five or six categories. In a particular year, I can’t invest in all of them. So I know, this year I want to double category A, grow category B by 70%, and category C by 30%. The rest, 10 – 15% growth.”

This isn’t arbitrary — each target reflects category maturity, market opportunity, and resource allocation decisions for the year.

2. Build Channel-Category Matrix

Next is the channel-category planning matrix with departmental input. ​“Decide with your E‑Com, QuickComm, GT, modern trade, and institutional teams — how much growth they see in each category.”

Each channel head brings specific insights: E‑com might see 50% growth potential for Date Bites based on search trends, while GT might only project 20% due to shelf space constraints.

“Combination of how much channel is growing and how my category is placed in that channel tells me the overall growth.” For instance, if quick commerce is growing 100% and your category indexing well there, you might plan for 150% growth. But if modern trade is flat and your category is underpenetrated, maybe just 10%.

The output is a detailed matrix: six product categories across five channels, creating 30 growth projections that roll up to company targets. Each cell reflects both market opportunity and execution capability.

3. Define Clear Ownership Structure

Abhishek’s org structure has two layers of ownership. ​“Your sales org is channel head wise. Each channel has one head.”

But the real P&L ownership sits with the brand team. ​“Each category has a brand head. The brand head owns the entire sales part, GM, supply chain — everything. The eventual owner of everything is the brand manager.”

This dual structure ensures execution capability (channel heads) while maintaining category P&L accountability (brand managers).

4. Map Sales to Procurement Planning

Farmley leverages ~8 years of historical data for SKU-level planning. ​“If I want to sell hundred of X, I need to know when to procure at best rates, how much to store versus spot buy. We track when each product’s prices fluctuate across six categories and 250 products — deciding what to contract, what to store in godowns, what to buy spot.”

5. Cash Flow and Supply Chain Planning

The procurement plan feeds into supply chain planning with a longer horizon. ​“Sales and procurement are planned each year. Supply chain has to be planned for three years because you can’t build CapEx each year.”

The supply chain team then provides ​“CapEx requirements, space requirements, fulfillment center requirements” based on category priorities.

6. Set Marketing Strategy by Category Maturity

The final piece is GTM planning, which varies by category age and growth trajectory. ​“You decide for each category, how long it has been in the system. Very early categories growing really fast — how much you want to put in performance marketing, how much in brand marketing.”

Performance marketing is platform-based with specific percentage allocations. ​“Performance marketing is percentage-linked to sales by channel. I’ll invest 15% of sales on quick commerce and 10% on e‑commerce. So if I’m scaling Date Bites to 3X on Blinkit, the marketing budget scales proportionally.”

Brand marketing follows a different logic with a critical constraint. ​“Only two categories can be picked per year to do large-scale campaigns to get any good top-of-mind awareness. When customers already know your brand in that category, you even see a dip in performance marketing needs.”

This forces tough choices: ​“Which category to choose for that year, how much percentage you are putting, which year of journey your category is in — these decide how much brand marketing I’m doing. I’ll do 50 crores this year, but I’ll do 90% of that in Date Bites, and 10% saliency on the rest five categories.”

Where D2C AOPs Get Messy

Even the best-structured AOPs encounter recurring challenges. Here are four that Romita and Abhishek have learned to navigate through experience.

Discount vs. Marketing Budget Trade-offs

For Romita, discounting is a negotiation centered on conversion math.

“If 100 people visit at a 10% conversion rate — that’s 10 conversions. If the product is ₹100 and 10 people convert, at ₹50 more should convert. At ₹200, fewer will convert. What price point gives us the conversion we need? Every channel has a different answer.”

The key constraint: ​“Budgets are control functions, not revenue worked backwards. I give you X budget. You tell me how much revenue you can generate, which means maximizing conversion and AOV.”

This becomes real-time negotiation between founder and channel head.

“When you say we can only sell a ₹100 SKU at ₹90, I might say that 10% discount isn’t possible. I’ll give you ₹1 extra in marketing instead. Or I’ll say 10% is fine — actually, give 15% but then I’ll either reduce your marketing budget or increase your revenue target.”

The channel head must optimize within constraints: ​“Given a marketing budget and targets for conversion and AOV, at what ASP can you operate to hit X revenue?” This forces complete equation thinking rather than just requesting more discounts or marketing spend.

Procurement, CapEx, and Seasonality

While Romita wrestles with pricing dynamics, Abhishek faces a different challenge: commodity price volatility. In commodities, procurement is about timing the market. ​“The more planned your procurement, the better your margins. Prices fluctuate significantly. You can’t buy when you need — you buy when the time is right,” Abhishek explains.

Farmley uses eight years of historical data to navigate seasonal price swings: ​“We track when each product’s raw material prices fluctuate, when to source and store in godowns, how much to contract versus spot buy.”

The scale is massive: ​“Six categories, approximately 250 products. Product by product, we decide how much to procure, which to buy now, which to store, which to spot buy.”

Each SKU requires three decisions — when to buy, how much to store, when to spot buy — all cascading into cash requirements. ​“If I want to sell hundred units of X, I need clarity on when to procure at best rates, when cash is needed, how much storage space is required.”

Seasonal peaks amplify complexity. Diwali planning starts six months ahead with monthly procurement milestones. Miss the window when makhana prices are low, and margins evaporate.

The strategic challenge: while sales and procurement are planned annually, supply chain infrastructure requires a three-year horizon. ​“You can’t build CapEx each year.”

This creates a rolling cycle where today’s Diwali sales inform next year’s procurement timing and infrastructure requirements two years out.

The Buffer Philosophy: Planning for Opportunity

When quick commerce exploded, Abhishek discovered why perfect AOPs fail: they lack buffers.

When Farmley entered Blinkit early — with just 70 – 80 dark stores — the channel exploded. ​“We went aggressive on quick commerce, and 90% of our growth came from there. Marketing was cheap then. It’s really expensive now.”

Without buffers in cash, inventory, and capacity, they’d have missed this growth. The lesson crystallized: ​“Whatever sales we expected from quick commerce, we’d plan at least 35% ahead. If a category was doing X sales on quick commerce, we’d plan for X plus 35%. Same with marketing — if we needed Y percent, we’d budget Y plus 35%.”

This 35 – 40% buffer now extends across the business: ​“We maintain a 35 – 40% buffer in procurement and supply chain planning to handle the demand fluctuations this channel creates. We have cash flow buffers for sudden procurement, payroll buffers for emergency hires, and marketing buffers for channel opportunities.”

Managing Working Capital and Cash Flow

At Farmley, working capital planning confronts a simple reality: food prices swing wildly, demand spikes unpredictably, and procurement windows are narrow. Miss the window when makhana prices are low, and margins evaporate.

“You can only buy at the right price with capital upfront. If you buy late or can’t pay, you miss the best margin windows.” The solution: stay 45 – 50 days ahead of cash requirements.

Abhishek suggests systematically layering funding sources. ​“You have to plan how much can I take from my banking line, which is cheap, but sustain my cash flow requirements which is normal months, and then how much I take from the NBFC/​working capital partners, which spikes up my interest rate, but eventually, fills up my working capital requirements.”

The lesson came from a painful experience. ​“We learned planning for perfection is less valuable than having a controlled buffer. The cost of missing ₹5 crore in sales is far greater than paying ₹1 crore in added working capital float.”

Now Abhishek deliberately over-plans: ​“In August, if I need ₹40 crores, and have a ₹25 crore banking line available, I need to arrange ₹15 crores. But I’ll actually arrange ₹20 crores — taking a ₹5 crore hit in extra interest to ensure we never miss an opportunity.”

The Control Systems That Matter

From Hands-On to Systematic Oversight

Building an AOP is one thing. Making it happen is another. As Foxtale scaled, Romita shifted from hands-on management to systematic controls. She says, ​“I used to see even ₹10 invoices. Now I only see invoices above a certain amount. We generate 100-plus invoices daily — I can’t see them all.”

This creates a paradox every scaling founder faces: If you’re not seeing 100% of everything, how do you still know everything? The answer is systematic controls. As a CEO, you need to know without seeing — that’s where controllership comes in.

Romita suggests, ​“When I say controls, it’s something you’re inflexible about. There’s no negotiation. A P&L head can’t come and negotiate. Costs are control functions — not one rupee can move from here to there.”

The enforcement is precise. ​“You can exceed your RoAS target. But if you’re below it, I’ll cut your budget. I look at RoAS daily. If it’s falling, costs get cut immediately because without proper RoAS, I don’t care about revenue. That’s not high-quality revenue. RoAS is the quality of revenue you’re driving. That’s your KPI as a P&L head. Costs are control functions — they’re predictable for us. It’s never a question of costs overrunning.”

24X7 Marketing Controls in Practice

At Foxtale, marketing spend — their biggest cost center — gets monitored every single day with surgical precision.

“Every day, we track exact RoAS for every channel. When I say channel-wise, I don’t mean just platforms like Nykaa. Meta is one marketing channel, Google is another, influencers are separate — with Instagram and YouTube tracked independently. Even partnerships are different. What I spend on Nykaa banners is tracked separately from Nykaa influencers.”

The granularity goes deeper: ​“We calculate RoAS bottom-up at the source of spend. We break down Nykaa into banners versus influencers. Meta gets broken down to campaign and product level.”

Real-time action is built-in. ​“Today, my head of digital marketing doesn’t just report that we’re at 1.8X RoAS. They’ll message saying it was at 1.8X, so they cut down spends and now it’s at 2.1X.” The team has internalized controls to where corrective action happens automatically.

The technology backbone — ​“dashboards built in-house using Python macros and PowerBI, connected to Meta, Google, everything” — pulls data directly from sources, eliminating human error.

“Controls is a mindset, the DNA of the company,” she emphasizes. The bottoms-up granularity means the control system operates proactively rather than reactively.

Daily to Quarterly: The Review Cadence

Abhishek breaks down his review system — how it moves from daily tracking to quarterly planning. At its core are comparative checks across different time periods: ​“If I’m on the seventh, how do the first seven days of sales compare to the last seven days of last month and the same seven days last year?”

Each day also breaks down category performance: ​“Sales category-wise, out of my six categories, how much sales I’ve achieved, how much is planned this month, and percentage growth. A rough daily MIS is a must to know how will you eventually perform in this month.”

Weekly huddles focus on execution: ​“How much of the actionables did you move? It could be infra, buying, team building, or store changes. I huddle at 10 am every Monday with all department heads.”

Monthly reviews check both AOP alignment and team gaps: ​“How far are you from the AOP and what can you do to correct? What manpower constraints are we facing to reach our goal?”

Quarterly reviews force market reality checks and enable dramatic pivots: ​“Say, one Qcom platform isn’t opening more dark stores next quarter. But we anticipated it will open 200 stores this quarter and expected a 40% revenue jump. I’m pulling out from that one and channelize my marketing money into the other Qcom platforms because they are still growing aggressively.”

These control systems — from daily RoAS monitoring to quarterly pivots — represent a fundamental evolution in how both founders operate.

Every month-end in fiscal year 23, Romita would plan the next month based on instinct. Today, her team cuts Meta spending in real-time when RoAS drops below 1.8X. Abhishek learned to buffer 35% above projections after nearly missing quick commerce’s explosion. Both evolved from annual spreadsheets to daily control systems — tracking everything from Nykaa banner performance to makhana procurement windows. The lesson is clear: in D2C, your AOP isn’t a document you present. It’s a system you operate, every single day.