First order offers for D2C: pick the shape, size the test

The welcome offer is a pricing decision in a marketing costume

Most Indian D2C brands pick a first order offer once, at launch, usually 10 percent off with a code the founder chose in a hurry. Nobody revisits it. That code then quietly sets average order value, gross margin, prepaid mix, and the discount expectation of every cohort that follows. On a Rs 899 order at 62 percent gross margin, 10 percent off is roughly 14 percent of gross profit, spent mostly on people who were going to convert anyway.

Treat the offer as a pricing decision with a defined shape, a defined cost, and a review date. Not as a permanent feature of the website.

Four offer shapes and what each one does

• Flat rupee off, such as Rs 150 off above Rs 899. Cost per order is fixed and predictable. It pulls carts up to the threshold. It reads least generous on high ticket baskets.
• Percentage off. Reads generous, scales badly. The bigger the cart, the more you pay, which is the opposite of what you want.
• Free gift, such as a travel size unit at Rs 60 COGS. Highest perceived value per rupee of real cost, and it leaves the price list intact. Adds pick and pack time and a stockout risk on the gift SKU.
• Bundle or set upgrade. Strongest AOV lever, because it converts a price cut into an assortment choice. Slowest to set up, since it needs its own page and inventory logic.

Simple rule. If the constraint is margin, use gift or bundle. If the constraint is conversion rate on cold traffic, use flat rupee off with a threshold at about 1.15 times your current AOV. Never run more than two shapes at once. Stacking a welcome code on top of a site wide banner sale is exactly how brands arrive at a 34 percent blended discount without anyone deciding to.

Size the test so the answer means something

Offer tests fail because they get read after 60 orders and a hunch. At a 2.2 percent conversion rate and a realistic effect size of 8 to 12 percent, you need roughly 400 to 600 orders per arm before the result is worth acting on. For most brands that is 10 to 21 days of traffic. Plan the calendar around it rather than stopping the test the moment one arm looks ahead.

Three rules keep the read clean.

• One variable per test. Offer shape or threshold or depth. Not two of them.
• Keep branded search, email, and WhatsApp traffic out of the test. Those audiences arrive with intent and will convert through almost any offer, which flattens the difference you are trying to measure.
• Write the decision rule before launch. Deciding after you see the numbers means the loudest person in the room wins.

Judge on contribution margin per session, not conversion rate. An offer that converts 6 percent worse and delivers 11 percent more margin per session is the better offer. Conversion rate is the metric that makes discount look like genius.

The prepaid nudge is usually the better rupee

In most Indian D2C categories, 45 to 60 percent of orders still arrive as cash on delivery, and those orders carry a collection fee and a materially higher chance of coming back undelivered. A 5 percent prepaid discount, or a flat Rs 50 off on prepaid, typically shifts 10 to 15 points of order mix to prepaid inside a month.

That is often a better use of the same rupee than a deeper welcome discount, because it turns a discount into a delivered order rather than a parcel that travels twice. Test it as its own experiment. Running a new acquisition offer and a new prepaid incentive in the same fortnight makes both unreadable, and the team will spend the next quarter arguing about which one worked.

The offer calendar and the discount floor

Write down a floor before you plan anything: a maximum blended discount as a percentage of gross revenue. For a healthy Indian D2C brand that is usually 8 to 14 percent, depending on category and repeat rate. Every offer decision for the quarter then has to fit inside that ceiling, which forces choices instead of accumulation.

A workable quarter looks like this. Two high depth moments tied to real events, each capped at seven to ten days. A steady state offer for the rest of the period. And at least four weeks of full price trading, so you learn what demand looks like without a coupon attached to it. Brands that never trade at full price have no idea what their product is worth, and they discover it during the first month a paid channel gets expensive.

Review three numbers monthly: blended discount as a percentage of revenue, AOV, and first order contribution margin. If discount is climbing while AOV stays flat, the offer has stopped buying growth and started paying for orders you already had. That is the point to change the shape, not the depth.