You want to start a bakery brand because the numbers look easy and the stories look inspiring. Theobroma began in 2004 as one small counter in Colaba, Mumbai, run by two sisters. In 2025 ChrysCapital agreed to buy a 90% stake at roughly ₹2,410 crore, on revenue that grew from ₹107 crore in FY20 to about ₹451 crore in FY24. Bakingo started in 2016 with ₹2 lakh, ran cakes out of cloud kitchens, raised $16 million from Faering Capital, and built a business now quoted in the ₹200 crore range. Every one of these started small, with an oven and one product.
Here is what those stories hide. Bakery is the only D2C category where your product starts dying the day you make it. A serum lasts two years. A brownie, if you want it to taste like a brownie, gives you days. That single fact decides everything: what you can sell, who you can sell it to, and whether you can ship at all. So before the budget tiers and the FSSAI forms, you have to answer one question that most first-time bakery founders skip.
That question is the shelf-life fork, and this guide resolves it, along with the full roadmap: FSSAI compliance, the cloud kitchen model, packaging that survives an Indian courier van, unit economics, and the honest ladder to ₹5 lakh a month.
Bakery in India is a large, fast-growing, low-entry-cost category with one brutal constraint: perishability. India's bakery market crossed US$12 to 15 billion in 2025 and is growing near 9% a year. The first decision is the shelf-life fork: shelf-stable packaged goods (cookies, brownies, cakes in a jar, at 15 to 90 day shelf life) that you can ship anywhere, versus fresh local delivery (cakes, pastries) that only travels within a city. FSSAI is mandatory from day one, and a home or cloud kitchen just needs Basic Registration at ₹100/year until you cross ₹12 lakh turnover. AOV runs ₹300 to ₹800. Gross margins on baked goods are strong at 55 to 70%, but shipping and spoilage eat the shippable line and last-mile logistics eat the fresh line. ₹50,000 starts a home bakery on packaged cookies. ₹2 lakh sets up a small cloud kitchen. ₹5 lakh buys a proper kitchen, packaging and ads. ₹1 lakh a month in revenue pays ₹15,000 to ₹28,000. ₹5 lakh a month takes 800 to 1,200 orders and pays ₹70,000 to ₹1.2 lakh.
What the Indian bakery market really looks like in 2026
The size is real and the growth is real. India's bakery market was valued between US$12 and 15 billion in 2025, growing at roughly 8.7 to 9.5% a year toward US$30 billion by the mid-2030s, pushed by urban incomes, gifting occasions and premium artisanal demand in metros and tier 2 cities. Bread and biscuits dominate the mass end. Your slice is the premium, branded, gifting and treat end, and you need the honest numbers of that slice, not the headline.
AOV band: ₹300 to ₹800. A single dessert or a small cookie box lands at ₹300 to ₹450. A celebration cake or a gift hamper pushes ₹600 to ₹800 and beyond. Below ₹300 the shipping or delivery cost eats you alive. Above ₹800 you are asking a stranger to trust an unknown kitchen with an occasion, which takes reviews and photos to earn.
Margin band: 55 to 70% gross on the product. Flour, butter, chocolate and sugar are cheap relative to the price a good brownie commands. The problem is never the ingredient margin. It is what happens after the oven: cold-chain, spoilage, breakage in transit, and the ₹70 to ₹90 that last-mile delivery costs on a fresh order.
RTO and spoilage exposure: this is your real risk. Unlike apparel, a returned cake is not resellable, it is garbage. For shippable packaged goods, a failed delivery or a crushed box is a total loss of a perishable unit, so you push prepaid hard and pack for abuse. For fresh local delivery, RTO is low because most orders are prepaid gifts, but a late or melted delivery is a refund and a bad review. The discipline for prepaid conversion is the same playbook as in how to reduce RTO on COD orders.
The competition, honestly
The two proven paths are visible in the brands you already know. Theobroma grew a retail-plus-online business over 20 years to nearly 200 stores across 30 cities before the ₹2,410 crore exit. Bakingo took the opposite route, no retail, pure online cakes fulfilled from 100+ cloud kitchens, and grew to a ₹200 crore business on the back of same-day city delivery. One is a bakery you visit. The other is a bakery that never lets you in, because there is no shopfront, only kitchens.
In 2026 you are not competing with an empty market. Every city has fifty home bakers on Instagram and a Bakingo delivering in two hours. "Chocolate brownies" is not a brand, it is a Swiggy search with a hundred results. The wedge that works is narrow: a specific product done unusually well, for a specific occasion or audience. Eggless gifting for a religious community. Sugar-free desserts for diabetics. Cookies that survive a courier so you can sell them pan-India instead of one city. Pick the corner nobody else is defending.
The shelf-life fork: shippable packaged vs fresh local
This is the decision the whole business hangs on. Make it consciously before you buy a single ingredient.
If your product can be made shelf-stable for 15 to 90 days (cookies, brownies, dry cakes, cakes in a jar, cookie boxes, gifting hampers) → you can ship pan-India through normal couriers, sell on your own store and Amazon, and your market is the whole country. If your product must stay fresh and refrigerated (celebration cakes, cream pastries, cheesecakes) → you are a local delivery business, your market is one city per kitchen, and scale means opening more kitchens, not more courier routes. If you want national reach on a modest budget → start shelf-stable, always. If you want high AOV occasion orders and can commit to a city → go fresh local and master same-day delivery.
Most first-time founders default to fresh cakes because that is what "bakery" means to them, then find out their entire market is a 15 km radius and every order needs a delivery partner. The shippable route is slower to feel glamorous but far easier to scale from a spare room. Cookies and brownies made for transit, sealed and dated, are a real national D2C business at a fraction of the operational load.
What ₹50,000 to ₹5 lakh actually buys you in bakery
Your budget decides your kitchen, not your ambition. Here is what each tier realistically buys in 2026.
| Budget | What it buys | Kitchen | Route | What it must prove |
|---|---|---|---|---|
| ₹50,000 | Home kitchen with a decent oven and mixer (₹18,000 to ₹25,000), FSSAI Basic Registration (₹100), first batch of ingredients and transit packaging (₹8,000 to ₹12,000), simple store and phone shoots (₹5,000), a small ₹8,000 to ₹10,000 ad or Instagram push | Home | Shelf-stable packaged (cookies/brownies) | That strangers buy your product at ₹349+ and it survives shipping |
| ₹1 lakh | Better home setup, sturdier packaging tooling, a 6-week ad and content test, ingredients for repeat batches, basic label printing and food-grade boxes | Home, scaling | Shelf-stable packaged | 150+ orders in 60 days with breakage under 5% and CAC under ₹200 |
| ₹2 lakh | A small shared or rented cloud kitchen (rent ₹10,000 to ₹25,000/month per industry rates), commercial oven and refrigeration (₹40,000 to ₹80,000), FSSAI State License (₹2,000 to ₹5,000), packaging and ₹40,000 to ₹60,000 of ads | Cloud kitchen (small) | Shelf-stable, or fresh local in one city | Repeatable weekly volume and the first repeat customers |
| ₹5 lakh | A proper cloud kitchen fit-out (₹2.5 to 4 lakh total per setup guides), commercial equipment, staff for one or two hands, transit-grade custom packaging, ₹1.2 to 1.5 lakh of ads over 90 days, working capital for ingredient buffer | Cloud kitchen (full) | Shelf-stable at scale, or multi-day fresh delivery | ₹1 lakh+ months with 20%+ repeat, the base for the ₹5 lakh climb |
Notice the trap in the top tier. A ₹5 lakh cloud kitchen fit-out with rent and staff creates a fixed monthly cost of ₹60,000 to ₹1 lakh before you sell a single box. If demand is not proven, that fixed cost bleeds you dry in three months. This is why the ₹50,000 home tier exists: to prove people buy before you sign a lease. The logic behind starting cheap is the same as in starting an online business with ₹50,000.
How to make and fulfil: the cloud kitchen route
You have three production models, in rising order of cost and commitment. A home kitchen is where almost every bakery brand should start: minimal fixed cost, FSSAI Basic Registration, and enough capacity to prove demand. A cloud kitchen (a delivery-only commercial kitchen, no dine-in, no shopfront) is the scale model Bakingo used: you rent 100 to 600 sq ft, install commercial ovens and refrigeration, and fulfil online orders only. A third-party co-packer makes sense only for shelf-stable packaged goods at real volume, where a licensed food factory bakes and packs your cookies to your recipe under their FSSAI license.
Real setup numbers to plan against:
| Item | Home kitchen | Small cloud kitchen | Full cloud kitchen |
|---|---|---|---|
| Rent / month | ₹0 (own space) | ₹10,000 to ₹25,000 (shared) | ₹20,000 to ₹50,000 (metro, 300 to 600 sq ft) |
| Equipment (oven, mixer, chiller) | ₹18,000 to ₹40,000 | ₹60,000 to ₹1.5 lakh | ₹2 to 6 lakh |
| FSSAI | Basic, ₹100/year | State, ₹2,000 to ₹5,000/year | State or Central, ₹2,000 to ₹7,500/year |
| Per-unit cost (brownie / cookie box) | ₹40 to ₹90 | ₹40 to ₹90 | ₹35 to ₹80 at volume |
Rent and equipment figures are drawn from published cloud kitchen cost breakdowns, where a small setup runs ₹2.5 to 10 lakh depending on city and scale. Whichever model you pick, packaging for transit is not optional and not cheap. A brownie that tastes perfect and arrives crushed is a refund. Budget ₹15 to ₹40 per unit for food-grade, cushioned, sealed, dated packaging on shippable products, and test it by couriering a box to yourself before you sell one. The sourcing method for ingredients and packaging in bulk is in how to find manufacturers and suppliers in India.
Founder Decision Loop™: signal, smallest honest test, hard read of the numbers, then commit capital. Applied to bakery: the signal is a specific occasion or audience with a specific craving, the smallest honest test is 30 to 50 boxes baked in your home kitchen and shipped, the hard read is breakage rate, repeat orders and CAC after 60 days, and the capital commitment is the cloud kitchen lease. According to the Founder Decision Loop™, demand validation comes before kitchen selection, because a beautiful cloud kitchen baking a product nobody reorders is just an expensive way to lose money slowly.
Compliance: what a bakery brand owner actually needs
Food is regulated tighter than most D2C categories, and skipping this gets your listings pulled and your kitchen fined. The good news is the entry bar is low and cheap.
- FSSAI, mandatory from day one. Every food business needs it, no exceptions. Which tier depends on turnover, per the FoSCoS portal: Basic Registration (₹100/year) if turnover is under ₹12 lakh, State License (₹2,000 to ₹5,000/year) from ₹12 lakh to ₹20 crore, and Central License (₹7,500/year) above ₹20 crore or if you operate across states. A home baker starts on Basic and never touches State until real volume arrives.
- GST registration. Mandatory to sell on any marketplace, regardless of turnover. Most baked goods sit in the 5% or 18% slab depending on packaging and type, so confirm your exact HSN with your accountant.
- Trademark. File in Class 30 (baked goods, confectionery) before you print a single box. ₹4,500 government fee for individuals and small enterprises, plus an agent fee if you use one. A brand name you cannot own is a liability the day you grow.
- Legal Metrology labels on packaged goods. Any sealed, packaged product must declare your brand and address, net weight, MRP inclusive of taxes, date of manufacture and best-before date, batch number, full ingredient list, the veg/non-veg mark, allergen declaration (eggs, nuts, milk, gluten), and consumer care contact. The best-before date is not decoration, it is the legal and food-safety spine of a perishable product.
- Hygiene and food safety. FSSAI expects safe handling, clean storage and correct labelling. For a home kitchen this is basic discipline. For a cloud kitchen it is documented practice, because inspections happen.
Budget ₹8,000 to ₹15,000 and one to two weeks for the full stack at the home tier. It is the cheapest insurance in this business, and unlike a factory license, you can genuinely do the FSSAI Basic Registration yourself online for ₹100.
Bakery unit economics: a ₹499 brownie box, line by line
Run every product through the Margin Waterfall™ before you commit to a kitchen or a batch size. According to the Margin Waterfall™ framework, contribution margin is calculated before the ad budget is set, not found out after the ads have spent it.
Margin Waterfall™: selling price minus COGS, packaging, shipping, payment gateway, spoilage and RTO loss, then CAC. If the number at the bottom is negative, no amount of scale saves it. In bakery the product margin is generous, but transit packaging and spoilage are heavier lines than in any other category, and that is exactly where the waterfall springs its leak.
Read that table like an operator. ₹104 on a ₹499 sale is a 21% net contribution, and it is fragile. Notice the two lines that don't exist in other categories: packaging at ₹35 and spoilage at ₹30 together take ₹65, more than the ingredients. Three levers protect the number:
- AOV. A ₹799 gift hamper ships in the same box as a ₹499 order but carries far more contribution. Gifting and bundles are the cheapest CAC hedge in bakery, because the occasion justifies the spend.
- Repeat rate. A brownie box gets finished in a week. The second order has near-zero CAC, so a strong repeat rate on a genuinely good product doubles blended profit per customer. Taste is the growth engine here in a way it never is in skincare.
- Prepaid share. Every COD order you convert to prepaid removes the risk of a perishable box coming back as pure loss. On food, COD RTO is not a discount, it is destroyed inventory.
Price with the waterfall, not with the home baker down the street. The complete method is in how to price a product in India.
In my supply chain and operations years, spoilage was the number I hated most, because it is loss you cannot sell your way out of. Bakery founders meet the harshest version of it. A box of cookies unsold on day 40 of a 45-day shelf life is not a markdown opportunity, it is a write-off with a smell. When I coach a bakery founder, I make them bake to confirmed orders first and to forecast second, never the reverse. Every batch you bake on hope instead of demand is a batch you are gambling against a clock that always wins. Start with pre-orders, weekly drops and made-to-order, and only build inventory once the reorder data proves the demand is real.
Where to sell bakery: own store vs Amazon vs Swiggy/Zomato
The channel call depends entirely on your side of the shelf-life fork.
| Platform | What it gives a bakery brand | What it costs you | Use it when |
|---|---|---|---|
| Your own store (Shopify or equivalent) | Full margin, customer data, pre-orders, gifting flows, repeat reminders | You buy every visitor with ads or content | Always, from day one, for both shippable and fresh. Repeat and gifting are the model, and only your store lets you own them |
| Amazon | National search demand for gifting and packaged treats, trust for unknown brands, prepaid buyers | 25 to 35% of MRP in fees, no customer data, strict food-listing rules | For shelf-stable packaged goods only. Cookies and brownie boxes with a real shelf life can sell pan-India here. Fresh cakes cannot |
| Swiggy / Zomato | Instant local demand, discovery, same-day delivery infrastructure | High commissions (often 20 to 30%), price-led buyers, thin margins on discounting | For fresh local delivery. This is how a cloud kitchen fills its city radius, the way Bakingo scaled. Not for national packaged goods |
The pattern that works: own store as the home base for both models, then Amazon if you are shippable and packaged, or Swiggy and Zomato if you are fresh and local. Do not try to be on all four in month one. Store build details are in the Shopify store setup guide for India.
The revenue ladder: what ₹1 lakh and ₹5 lakh a month actually take
Revenue targets without order math are wishful thinking. Here is the ladder at bakery's real numbers, profit shown beside revenue because in a perishable category revenue that comes with spoilage is not profit.
| Stage | Orders / month | AOV | What it takes | Owner's profit / month |
|---|---|---|---|---|
| ₹30,000 / month | 70 to 90 | ₹399 | Home kitchen, one hero product, Instagram plus word of mouth, made-to-order to avoid spoilage | ₹5,000 to ₹9,000 |
| ₹1 lakh / month | ~220 | ₹449 | 2 to 3 products, one working ad angle, breakage under 5%, prepaid share 50%+, first repeat customers | ₹15,000 to ₹28,000 |
| ₹3 lakh / month | ~550 | ₹549 | Small cloud kitchen, gifting bundles lifting AOV, 25% repeat rate, Amazon or Swiggy live alongside the store | ₹45,000 to ₹75,000 |
| ₹5 lakh / month | 800 to 1,200 | ₹499 to ₹649 | Full cloud kitchen, staff, 30%+ repeat rate, festival and gifting spikes, ₹1.2 to 1.8 lakh/month ad spend, tight ingredient forecasting | ₹70,000 to ₹1.2 lakh |
Two things about the top rung. First, the jump from ₹1 lakh to ₹5 lakh is not "more ads," it is repeat rate and gifting occasions. A genuinely good brownie brings the customer back without a second ad rupee, and festivals and birthdays deliver predictable spikes you can plan inventory around. Second, staffing and fixed cost arrive here whether you like it or not: at 1,000 orders a month you cannot bake alone, so a chunk of that revenue goes to hands and rent before it reaches you. This is why the profit line stays modest as a percentage even as revenue climbs. The stage-by-stage execution detail lives in the roadmap to ₹5 lakh a month.
Realistic timeline: what 30 days and 90 days actually look like
Days 1 to 30 (home kitchen, shippable): pick the product and the occasion, get FSSAI Basic Registration (₹100, online), perfect one recipe, source ingredients and transit packaging, courier a test box to yourself and a friend in another city, set up a simple store, shoot on a phone, and take your first pre-orders. A shippable bakery brand can genuinely be live and selling by day 30.
Days 1 to 90 (cloud kitchen or fresh local): weeks 1 to 3 to lock recipes and validate demand from the home kitchen first, weeks 3 to 6 to find and fit out a cloud kitchen and get the FSSAI State License, weeks 6 to 9 for equipment, staff trial and packaging, weeks 9 to 13 for launch and the first ad and delivery-platform experiments. Anyone promising a fully staffed cloud kitchen brand in 30 days has not waited on a commercial oven install or an FSSAI inspection. The day-by-day version is the 90-day D2C launch roadmap.
Before either clock starts, run the validation gate. In bakery this is doubly important, because you cannot warehouse your way out of a bad guess.
Validation Sprint™: a fixed-budget, fixed-deadline test that buys evidence instead of a lease. For bakery: bake 30 to 50 units in your home kitchen, take pre-orders through ads or your own network, ship or deliver them, and read the result after 14 days against pre-written pass/fail numbers: reorder intent above 40%, breakage under 5% on shipped orders, and CAC under ₹200. Pass, and you sign the kitchen lease with confidence. Fail, and the product or the occasion changes before the fixed cost does.
The full method for reading a test honestly, including what counts as a false positive, is in how to validate a business idea.
The mistakes that kill first bakery brands
Signing a cloud kitchen lease before proving demand. A first-time founder takes ₹4 lakh, rents a 400 sq ft kitchen at ₹25,000/month, buys ₹1.5 lakh of commercial equipment, and hires one helper, all because a "real bakery" feels serious. Now there is a fixed cost of ₹60,000 to ₹80,000 a month before a single order. Sales trickle in at 40 to 60 a month because the product was never validated, spoilage eats the rest, and by month three the founder is baking against a lease they cannot cover. Loss: ₹3 lakh and a shut-down, versus the ₹15,000 home-kitchen Validation Sprint™ that would have proven or killed the idea first. In bakery, fixed cost is the killer, and it is always signed too early.
The other repeat offenders, shorter: making fresh cakes and assuming you can ship them nationally, then watching them arrive melted; ignoring transit packaging and refunding half your first shippable orders; baking to inventory on hope and writing off spoilage every week; pricing at ₹199 to look affordable and finding shipping and spoilage ate the whole margin; skipping the veg/non-veg and allergen labelling and getting listings pulled; and treating Swiggy and Zomato discounting as growth while the commissions and price wars quietly delete your profit.
Execution checklist
- Make the shelf-life fork decision first: shippable packaged (national) or fresh local (one city). Write it down.
- Write your wedge in one sentence: which product, for which occasion or audience, done unusually well.
- Get FSSAI Basic Registration (₹100, yourself, online) before you sell one box.
- Perfect one hero recipe, then courier a packed box to yourself in another city and see what arrives.
- Run a Validation Sprint™ from your home kitchen with pass/fail numbers written before the test.
- Budget ₹15 to ₹40/unit for cushioned, sealed, dated transit packaging on shippable goods.
- Run the ₹499 Margin Waterfall™ on your own numbers; kill any product that dies on the packaging and spoilage lines.
- File the trademark in Class 30 and register GST before printing branded boxes.
- Build labels against the food declaration list: brand, net weight, MRP, best-before date, batch, ingredients, veg/non-veg mark, allergens, consumer care.
- Bake to confirmed orders and pre-orders first; only build inventory once reorder data proves demand. Never sign a kitchen lease before that.
Your next action
Today, do one thing: decide your shelf-life fork and bake one test batch of your hero product this week. If it is shippable, pack a box and courier it to a friend in another city to see what survives. If it is fresh, deliver a few orders across your city and time them. That single batch turns this whole guide from reading into a real decision on your own product, your own kitchen, your own numbers. Everything else, the FSSAI form, the store, the cloud kitchen, sequences behind that one batch. The founder frameworks referenced through this guide come from Ravikant Tyagi's operating system for exactly this journey.
If you'd like the complete execution system, calculators, SOPs, templates and operating frameworks behind this process, continue inside D2C Acquisition.Lab.
