Ask most D2C founders who their product is for, and the answer is "everyone."
It sounds like ambition. In the ad account, it behaves like a tax.
The fear behind it is fair. Going niche feels like shrinking your market before you've sold a single unit. But brands rarely struggle because they went too narrow. They struggle because they never chose at all, or because they chose a niche so small that even owning it couldn't pay the bills.
Both mistakes are avoidable, and both can be caught on paper before you spend on stock or ads. This is the framework we'd use: five steps, a sizing sheet, and a scorecard at the end so you can test your niche honestly.
The framework at a glance
- 01
Define the niche as a person, a problem and a moment.
Not a product.
- 02
Pick your angle.
Own who it's for, or own what it's best at. One, not both.
- 03
Size it before you commit.
If 5% of the niche can't get you to ₹10 lakh a month, widen one step.
- 04
Launch one hero product.
Put all your money, reviews and story behind one thing.
- 05
Widen one step at a time.
Same customer, next need.
First, why "for everyone" costs you money on Meta
Most new D2C brands now run broad targeting on Meta: no interests, no age filters, just the country and a budget. That's the right setup. But it changes where your niche lives.
When you don't tell Meta who to target, Meta works it out from two things: your creative, and the people who respond to it. It shows your ad to a wide audience, watches who stops, who clicks and who buys, and then goes looking for more people like them.
Now follow what happens to an ad written for everyone.
"Premium quality. Best price. Made for you."
Nobody reads that and thinks this is for me, so nobody in particular stops. The few clicks you get come from a random mix of people with nothing in common. Meta has no pattern to learn from, so it keeps showing the ad to more random people. You keep paying for impressions that scroll past, and your cost per purchase climbs.
Now change one line.
"For working women who hate ironing their kurtas every morning."
The right woman stops, because the ad is plainly talking to her. She clicks. Some of the women like her buy. And Meta now has a clear pattern: it goes and finds more of her.
Ad written for everyone
- First line
- "Premium quality. Best price."
- Who stops
- Nobody in particular
- What Meta learns
- Nothing useful
- What happens to cost
- Rises as Meta keeps guessing
Ad written for a niche
- First line
- "For working women who hate ironing their kurtas every morning."
- Who stops
- The exact customer you want
- What Meta learns
- A clear profile of who buys
- What happens to cost
- Falls as Meta finds more of the same buyer
Step 1: Define the niche as a person, a problem and a moment
A category is not a niche. "Kurtas" is not a niche. "Razors" is not a niche. "Skincare" definitely isn't.
A niche has three parts:
01 · Who
A person
Who exactly is buying?
02 · What
A problem
What's going wrong for them today?
03 · When
A moment
When do they feel that problem most?
| Category | Not a niche | A niche |
|---|---|---|
| Kurtas | "Kurtas for women" | Working women who need office kurtas that don't need ironing every morning |
| Razors | "Razors for men" | Men who find shaving a chore and put it off until they have to |
| Jeans | "Jeans for every woman" | Curvy women who have never found jeans that fit at both the waist and the hips |
| Baby care | "Baby products" | New parents who read every ingredient label before buying anything for their child |
Each part does a job. The person tells you who to cast in your ads. The problem gives you your hook. And the moment gives you the scene: the morning rush before office, the wedding season, the first humid week of summer. Ads that show the moment feel familiar before the viewer has even read the text.
If you've already written your product's problem as one line (the first step in how to pick your first D2C product), you have most of this. The niche is that line, made specific enough that a stranger can tell in one read whether it's about them.
Step 2: Pick your angle — own who it's for, or own what it's best at
There are two ways to narrow, and both work.
Route 1
Own an audience
You build for a specific person. Jeans made for curvy women. Skincare for oily skin in humid cities. Kurtas for women who commute to an office. The product may suit other people too, but the brand speaks to one.
Route 2
Own an attribute
You build around the one thing you do better than anyone. The best fabric. The best fit. The best design. The finest embroidery. Anyone can buy from you, but everyone knows what you're for.
What doesn't work is claiming everything at once. "Best fabric, best fit, best design, best price" is "for everyone" again, just with more words. Pick one thing and own it.
A useful way to separate the two ideas: your product can be for many people. Your positioning can't. Plenty of women who aren't curvy might happily buy a brand's jeans. That's fine. The positioning still has to pick someone to talk to, or it talks to nobody.
How to choose between them
Use review mining: a sheet of the repeated complaints in the 1–3 star reviews of the top sellers in your category.
- If the complaints cluster around a type of person ("never fits if you're petite", "useless for sensitive skin"), that's an audience route.
- If they cluster around a quality of the product ("fabric pills after a month", "embroidery comes loose"), that's an attribute route.
Either way, the market is telling you where the gap is. You just have to listen to the most repeated complaint rather than the most exciting idea.
Look for the niches big brands serve badly
Large brands build for the middle of the market, because that's where their volume is. The edges are where they're slowest: specific body types, specific skin or hair concerns, specific regional crafts, specific diets. These edges are often too small for a large company to prioritise, and more than big enough for a new D2C brand. That's not a weakness in your niche. It's your opening.
Step 3: Size it before you commit
This is the step most founders skip, and it answers the real fear: is my niche too small?
You don't need a research report. You need one line of maths.
Then take 5% of that. Five percent is a deliberately modest share, the kind of slice a focused brand can realistically win over a few years. It's a reality check, not a forecast.
A useful bar is ₹10 lakh a month, which is ₹1.2 crore a year. So 5% of your niche needs to be at least ₹1.2 crore a year. Put the other way: your niche should be worth at least ₹24 crore a year.
₹10 lakh
The monthly revenue bar
₹1.2 crore
The same bar, per year
₹24 crore+
The yearly niche size at which 5% clears it
The sizing sheet
| Niche | Reachable buyers | Spend per buyer, per year | Niche size | Your 5% share | Per month | Verdict |
|---|---|---|---|---|---|---|
| Working women in metro cities buying office kurtas | 20,00,000 | ₹3,000 | ₹600 Cr | ₹30 Cr | ₹2.5 Cr | Plenty of room |
| Curvy women who struggle to find jeans that fit | 8,00,000 | ₹2,500 | ₹200 Cr | ₹10 Cr | ~₹83 L | Room to grow well past ₹10L/month |
| A very narrow niche | 50,000 | ₹2,000 | ₹10 Cr | ₹50 L | ~₹4 L | Too small. Widen one step |
Look at the first row. A niche that sounds small, office kurtas for working women, can carry a brand many times bigger than ₹10 lakh a month. Niche doesn't mean small. It means specific.
Now look at the last row. Fifty thousand people spending ₹2,000 a year gives you about ₹4 lakh a month even if you win your full 5%. That isn't a business yet. It's a hobby with a website.
Where the numbers come from
You're after the right order of magnitude, not a precise figure. Three free sources get you close enough:
Source 1
Meta's audience estimate
Create a draft ad set in Ads Manager, set the location, age and gender that match your niche, add a couple of related interests, and read the estimated audience size on the right. It shows a range. It counts people Meta can reach, not people who'll buy, so treat it as a ceiling, not a target.
Source 2
Marketplace sub-categories
Search the niche, not just the category: "jeans for curvy women", not "jeans". If listings for the niche carry hundreds or thousands of ratings, the niche is already spending money.
Source 3
Your own conversations
Ask people in the niche how often they bought in the category last year and what they paid. That gives you the spend-per-year number from real behaviour instead of a guess.
Then be conservative. If your first estimate says 10 lakh buyers, use 5 lakh. If the niche still clears ₹24 crore a year, you're on solid ground.
If it's too small, widen one step
When a niche fails the sizing test, the answer is not to go back to "everyone." It's to take one step wider and test again.
Too narrow
- "Curvy women in Bengaluru who want office jeans"
- "Vegan gym-goers who want a protein bar with no added sugar"
One step wider
- "Curvy women who want jeans that fit"
- "Gym-goers who want a protein bar with no added sugar"
Keep the problem. Loosen one filter at a time. Stop at the first version that clears the bar.
Step 4: Launch one hero product
Once the niche is clear, the next temptation is to launch with a full range so the brand looks established. It almost always backfires.
Let's say you have ₹5 lakh to launch.
What ₹5 lakh buys you (illustrative numbers)
20 products
- Stock budget per product
- ~₹25,000 each, a tiny batch of everything
- Ad testing
- Budget split twenty ways; no product gets enough data
- Reviews after three months
- Twenty pages with four reviews each
- What customers remember
- Not sure what the brand is
- Dead-stock risk
- High: some of the twenty won't sell
1 hero product
- Stock budget per product
- The full budget behind one product, with the variants it actually needs
- Ad testing
- Every rupee teaches you about the same product
- Reviews after three months
- One page with dozens of reviews
- What customers remember
- "Oh, they're the ___ people"
- Dead-stock risk
- Low: you learn fast and reorder what works
A hero product concentrates everything a young brand is short of: money, data, reviews and attention. It also makes the brand easy to describe, which is what word of mouth runs on.
This is the pattern behind many of India's best-known D2C brands. Bombay Shaving Company entered a market where one global brand held close to 90% share, and it didn't try to match the range. It launched with one precision razor, sold online. Skin, hair, bath and women's grooming came much later. Mamaearth began with a handful of baby-care products for parents who wanted toxin-free options, and only moved into adult skin and hair once those parents asked for the same thing for themselves. Pee Safe started with a single toilet-seat sanitiser before building a full hygiene range.
None of them launched as a range. They launched as an answer to one problem.
Step 5: Widen one step at a time
Narrow to launch, widen to scale. The order matters, and so does the direction.
When to widen
Don't widen because you're bored of the hero product or because a new idea feels exciting. Widen when the evidence says the customer wants more from you:
- The hero product sells steadily, month after month, without you having to push harder every month to keep it there.
- Customers start asking for the next thing on their own ("do you make this in a trouser?", "is there a version for adults?").
- Your best buyers are coming back, and you have nothing new to sell them.
How to widen
The safest step is the same customer, with their next need. You already know how to reach this person, what they care about and what they'll pay. Selling them a second product costs far less than finding a new customer.
| Starting hero | Good next step (same customer, next need) | Risky leap (new customer and new need) |
|---|---|---|
| Non-iron office kurtas for working women | Matching bottoms, then office co-ord sets | Men's formal shirts |
| Jeans for curvy women | Trousers, then dresses with the same fit logic | Kids' wear |
| Baby wash for label-reading parents | Baby lotion and oil, then the same clean range for parents | Pet grooming |
A leap isn't impossible. But stepping into a new customer and a new need at the same time is effectively starting a second brand, funded by the first one. Do that only once the first brand can carry it.
The niche scorecard
Score your niche against these six questions. Be strict: give the 2 only when you have the evidence, not when you believe it.
| Question | 0 | 1 | 2 |
|---|---|---|---|
| 1. Definition: can you write it as a person + problem + moment in one line? | Only a category | A person or a problem, not both | All three, clearly |
| 2. The "that's me" test: would the right person know it's for them from your ad's first line? | Could be any brand's ad | Somewhat | Instantly, and someone outside the niche would shrug |
| 3. Size: does 5% of the niche reach ₹10L a month (a niche worth ₹24 Cr+ a year)? | Well short | Close, on optimistic numbers | Clears it on conservative numbers |
| 4. Spending proof: is the niche already paying for a solution? | No sign of spending | Some niche listings or searches | Niche listings with hundreds or thousands of ratings |
| 5. Room to own: are big brands serving this niche badly or not at all? | Well served already | Served, but with clear complaints | Largely ignored, or served badly |
| 6. Widening path: is there an obvious next need for the same customer? | None | One, loosely related | Two or three clear next steps |
Total out of 12
10–12
Go
Build the hero product and write every ad for this one person.
7–9
Sharpen first
A weak "that's me" score is usually a wording problem you can fix today. A weak size score means widening one step.
Under 7
Rethink
Keep the research. The right niche is often one step to the side of this one.
What to do this week
Here's what to do:
- Write your niche as one line: person + problem + moment. Write three versions if you're unsure.
- For each, choose your angle: own the audience, or own the attribute. Only one.
- Size each version: reachable buyers × spend per year. Check that 5% clears ₹1.2 crore a year. If it doesn't, widen one step and size it again.
- Write the first line of your ad for the winner. If it could sit on any competitor's ad, sharpen it.
- Choose your hero product and list only the variants it truly needs.
- Sketch the next two steps: what does this same customer need after the hero?
- Score the niche out of 12.
If your niche clears the scorecard, you'll go into your first ads with something most new brands don't have: a clear answer to "who is this for?" And that answer will do more for your cost per purchase than any targeting setting.
Picking the niche is one of the earliest decisions you'll make, and it shapes almost everything after it: your product, your creative, your price and who Meta finds for you.
Already doing ₹10L+ a month? Then the niche question comes back in a new form: which audience or product line to widen into next without pushing your acquisition cost up. That's the work we do, and the free audit below is where it starts.