Selling chocolate D2C in India: launching into a crowded shelf

Chocolate is a wonderful thing to make and a difficult thing to sell online in India. The shelf is dense, the claims are near identical, and the buyer cannot taste anything through a phone. Clean label, single origin, tree to bar, no refined sugar, 70 percent, 85 percent: the vocabulary is shared by everyone in the category, which means it differentiates nobody in it.

We have built three chocolate brands through that shelf and the lessons repeat. None of them won on recipe. They won on having something specific to say, finding out by testing which bars actually sell to strangers, and refusing to buy volume with discounts.

Your recipe is not your differentiator, your point of view is

Soothys launched with no digital presence, no past campaigns, no benchmarks, and a website converting at 1.3%. Product-market fit was a hypothesis. The category offered no cover, because clean-label chocolate is full of brands making near-identical claims about sugar, sourcing and ingredients.

The first correction was not a targeting change. It was one line the brand agreed to defend: guilt-free indulgence that actually tastes like a treat. That single decision settles a hundred smaller ones, because which SKU leads, which photograph runs and what the landing page says above the fold now all have a test to pass. Three months in, Soothys had grown revenue 208%, order volume had doubled and conversion had gone from 1.3% to 6%.

The practical version for you: write two or three competing propositions, not variations of one, and run them against each other as ads. The output should be a decision, not a preference. In a saturated food category, every competitor buys the same audiences with the same ingredient claims, so targeting is rarely where the advantage sits.

Let testing choose the hero bars

Founders know which bar is their favourite. That is almost never the bar that converts cold. On Soothys, structured testing across SKUs, audiences, formats and hooks surfaced two hero products that now drive over 60% of sales, which changed the whole budget logic: the catalogue stopped being treated equally and the proven sellers carried the spend.

Run this deliberately. Give each candidate bar enough budget to produce a readable result rather than spreading thin to look busy, read them separately, then let the winners anchor the account. The rest of the range is not dead, it just earns its money after the first order rather than fighting for the first click.

Structure the account so the hook and the close are read separately

On Soothys, testing on Meta was structured across awareness, engagement and conversion rather than run as one flat pool. That separation matters because the hook that wins attention is rarely the asset that closes a sale, and mixing them hides which one is failing. Cheap conversions from warm audiences flatter expensive ones from cold, and the account reports a number nobody can act on.

The same discipline appears on Anuttama, an 85% single-origin artisanal chocolate brand that wanted more revenue without the slide in efficiency that usually comes with it. Keeping cold prospecting, warm retargeting and existing buyers in separate structures stops the account paying prospecting prices for people already close to converting.

Growth and efficiency are not a trade-off if you add surface area

Anuttama's revenue climbed 200%, tripling, while ROAS doubled over the same period. That is unusual and worth understanding, because the normal pattern is the opposite: push more budget through a narrow base of a few creatives and one reliable offer, and the auction charges more for the same people while frequency climbs.

Three things made volume and efficiency move together. The creative pool was expanded first, so scaling did not mean fatigue and frequency spread across assets instead of stacking on one. Audiences were segmented rather than merged. And the retention layer was reinforced, so every new customer added compounding value rather than one-off revenue. Budget then rose only on the ad sets that kept their return as they took more volume, and stayed flat on the ones that did not.

Gifting is a second business inside the first

Bon Fiction makes tree-to-bar craft chocolate and had been burned enough times to believe performance marketing simply did not work for them. Online sales were a trickle, mostly limping through WhatsApp orders. The strategy that turned it around leaned into what makes the brand special: premium, narrative-led campaigns built around indulgence, gifting and craft, rather than price. Targeting aimed at tier-1 cities and affluent non-metros, which unlocked five new performing states, because demand for premium gifting outside the metros is routinely underestimated and usually cheaper to reach.

Treat gifting as its own plan rather than a seasonal banner:

• Build gifting SKUs on purpose. A box that looks like a gift converts at a different price to the same chocolate in everyday packaging.
• Separate the campaigns. The gifting buyer is not the buyer of your everyday bar, and often is not the eater at all, so the message is about the recipient.
• Work backwards from the delivery date, not the festival date. Cut-offs, courier lead times and a visible delivery promise decide whether the sale happens.
• Plan the corporate gifting enquiry route. One well-placed landing page and a form can carry more margin than a month of retail bars.
• Expect the season to end. Build the everyday reason to buy before Diwali, or you rent customers once a year.

Why discounting damages a craft brand specifically

Discounting is the fastest way to make a quarter look good and the fastest way to dismantle a premium chocolate brand. The argument you are making is that your bar costs more because of the cocoa, the process and the people. A recurring price cut is a public admission that it does not, and buyers learn the sale calendar within two cycles.

Bon Fiction's growth came from premium narrative rather than price, and it grew over 6× in five months entirely on paid media, with no cash-on-delivery crutch propping up the numbers. Prepaid revenue is real revenue, without the return and cancellation leakage that inflates a COD-heavy top line.

Where you do need a commercial lever, use one that raises basket value instead of lowering the anchor: a bundle-only saving, a discovery pack, free shipping over a threshold, or a limited edition. Anuttama's own read of its buyers says the same thing: buyers were happy to spend more per order when the value was framed clearly.

Mind the label and the claims

Clean-label chocolate lives close to the line on nutrition claims. Under the Food Safety and Standards (Advertising and Claims) Regulations, 2018, claims have to be truthful, substantiated and consistent with what is on the pack, and claims about reducing disease risk that are not listed in the Schedule need pre-approval from FSSAI. Section 53 of the FSS Act carries a penalty of up to ₹10 lakh for misleading food advertisements. Cyril Amarchand Mangaldas has a readable summary, and the regulations sit on the FSSAI regulations page. If your creative says "no added sugar" or "sugar free", make sure the pack says the same thing and that you can evidence it.

If performance marketing "does not work" for you

Bon Fiction's founders had concluded the channel was wrong for the brand. It was not. The account had broken tracking and attribution nobody could trust, so every optimisation decision was a coin toss and the account got worse the longer it ran. The fix was unglamorous: repair the tracking, decide who exactly to target and which SKUs deserve the spotlight, rebuild campaign objectives so the platform optimises toward the outcome the business wants, and split the funnel into stages measured separately. Google Ads, completely untapped until then, became a top sales driver.

In a saturated category, a clear point of view does more work than a better product. Soothys did not need a new recipe. It needed a reason to be chosen.

The short version

• Pick one proposition you will defend and test it against rivals, not against variations of itself.
• Find your two hero bars with money and data, then let them carry the account.
• Read prospecting, retargeting and repeat buyers as separate numbers.
• Add creative variety and audience segments before you add budget.
• Build gifting as a separate plan with its own SKUs, messages and delivery promises.
• Raise basket value instead of cutting the anchor price.

The media and creative side of this is performance marketing and creative and content, and the positioning work sits in strategy. More of the category is under food and beverage and FMCG.