The returns policy: written for the buyer's courage, priced for your margin

The policy is marketing: what it does before anyone returns anything

The overwhelming majority of people who read your returns policy never return anything — they read it to decide whether to buy, and whether to risk prepaid payment over COD. That makes the policy a conversion document first: its job is to remove the fear of being stuck with a bad purchase from an unknown brand. The evidence shows up in the funnel: clear, humane return terms near the buy button lift conversion and shift payment mix toward prepaid — which, per the RTO arithmetic, is one of the most valuable behavioural shifts available to an Indian D2C brand. Write the policy for the nervous first-time buyer, then engineer the edge cases separately.

What Indian buyers actually need to see

• The window, in plain numbers — "7-day return, 48-hour issue reporting for perishables" beats paragraphs. State when the clock starts (delivery date, not order date).
• Who pays return shipping — say it explicitly for each case: your fault (damaged, wrong, expired) versus change of mind.
• Refund speed and method — "refund to source within 5–7 days of pickup; instant store credit option" answers the real anxiety, which is not "can I return" but "will I chase my money for a month".
• The unboxing-video question — if you require one for damage claims (many Indian brands reasonably do), say so upfront, politely, at delivery time via WhatsApp — not as a gotcha after the claim.
• Category honesty — food, supplements and personal care cannot take opened-product returns; explain the why (safety), and offer the alternative (replacement or credit for quality issues with a photo).

A generous-sounding policy with sharp edges converts better and costs less than a stingy one with loopholes.

Food and consumables: the replacement-first model

For edible categories the standard return model breaks — you cannot restock a returned jar of ghee, so "return" really means "refund plus write-off plus reverse logistics on a product you'll destroy". The model that works: replacement-first, photo-verified, no return leg. Damaged in transit, quality issue, wrong item: customer sends a photo on WhatsApp within 48 hours, you ship a replacement or credit immediately, no parcel travels backwards. It costs you product COGS instead of COGS plus two shipping legs plus dispute friction — and it creates the service-recovery loyalty moment described in CX as branding. Reserve actual pickups for high-value non-consumable SKUs where recovery has real value.

Pricing the policy: the margin mathematics

Model it like an insurance product. Inputs: return/complaint rate by category (quality food brands typically run low single digits), average resolution cost per case under each policy design (refund vs replacement vs credit), fraud/abuse rate, and the conversion lift a friendlier policy buys. The maths almost always says the same thing: the cost of generosity on genuine cases is small, the conversion and repeat effects are large, and the real leak is not kindness — it is undesigned kindness: no photo verification, refunds before pickup on returnable goods, no abuse tracking, support improvising a different answer per customer. Fix the design, not the generosity. Store credit as a first-offered (never forced) option retains a surprising share of refunds as future revenue.

Abuse control without insulting the honest 97 percent

The guardrails that work quietly: photo/video verification for damage claims; a customer-level flag after repeated claims (second claim gets a human look, third gets manager approval); address- and pincode-level patterns watched monthly; COD limits or prepaid-only for flagged repeat abusers; and serial/batch codes on higher-value SKUs. What does not work: making every customer pay for the sins of a few with hostile policy language, demanding original-packaging perfection on a food pouch, or dragging refunds as a deterrent — the reviews that behaviour buys cost more than the fraud it prevents. The tone target: airline-lounge polite, casino-precise.

Operationalising it: the policy is a workflow, not a page

Publish the page (linked in the footer, the PDP trust strip and order emails), then build the machinery: WhatsApp as the claims channel with a guided flow (order number → issue type → photo → resolution offer), resolution SLAs (first response same day, decision within 24–48 hours), templated but human replies in the brand voice, refund triggers wired to pickup scans where a return leg exists, and a monthly review of claim themes — because a spike in "arrived damaged" is a packaging or courier-lane problem wearing a policy costume (the courier audit finds it). Done right, the returns experience becomes one of your highest-converting reviews: "had an issue, they fixed it in a day" is proof no ad can buy.