D2C Loyalty Points: Breakage, Liability and Real Cost

A loyalty programme is a discount you pay later

When a D2C brand launches points, the profit and loss looks unchanged for two quarters. Orders come in at full price, points accrue quietly, and the team congratulates itself on a retention win that cost nothing. Then redemption starts, and the discount that was deferred lands all at once, usually in the quarter when the brand is also funding a festive push.

The mistake is treating the earn rate as a marketing feature rather than as a liability with a schedule. Every point issued is a claim on future revenue. Your finance team will eventually book it as deferred revenue, but long before that it should sit in your contribution model as a real cost line with a forecast attached.

The four numbers that define the programme

Everything about loyalty economics comes down to four inputs. Get these on one page before you write a single line of programme copy.

• Earn rate: the percentage of order value returned as points. Three to five percent is the common Indian D2C band.
• Redemption rate: the share of issued points that eventually get spent. Typically 30 to 35 percent in year one, drifting to 45 to 55 percent by year three as balances age and habits form.
• Breakage: the mirror of redemption, the share that expires or is abandoned. It funds the programme, and it shrinks every year.
• Incremental repeat lift: the additional repeat rate among members versus comparable non members. This is the only number that justifies the other three, and it is the one almost nobody measures.

The effective cost of the programme is earn rate multiplied by redemption rate. A 5 percent earn rate with 38 percent redemption costs 1.9 percent of revenue. That is a genuinely cheap retention lever compared with a 10 percent welcome back code, which costs 10 percent on every order it touches. The catch is that the 1.9 percent is an average across all shoppers, while the actual cost concentrates on your most valuable cohort, the one that would have repurchased anyway.

Working the numbers on a Rs 4 crore brand

Take a brand doing Rs 4 crore of annual online revenue with a 55 percent gross margin. A 5 percent earn rate issues Rs 20 lakh of points a year. At 38 percent redemption the cash cost is Rs 7.6 lakh, or 1.9 percent of revenue, which is about 3.5 percent of gross profit. Outstanding liability at steady state, assuming a nine month rolling expiry, settles somewhere around Rs 6 to Rs 8 lakh. That is the number to put in front of the founder before launch, because it is the number that shows up as a cash claim during the exact quarter when inventory is being built for festive.

Now stress it. If redemption climbs to 55 percent by year three, on a business that has grown to Rs 8 crore, the cash cost is Rs 22 lakh. If the programme has not moved repeat rate by at least two to three percentage points by then, it has become an expensive rebate on loyal shoppers. Model that outcome on day one and set the review trigger in the calendar.

Design rules that protect contribution

Five rules do most of the work of keeping a points programme solvent.

• Cap redemption at 20 percent of cart value. This prevents fully paid orders, keeps average order value from collapsing on redemption days, and stops the programme from becoming a free product engine.
• Earn on prepaid only. In categories where cash on delivery is 40 to 55 percent of orders, tying points to prepaid typically moves prepaid share up by four to seven points, which pays for a chunk of the programme in reduced return to origin alone.
• Use a rolling expiry from last activity, not a hard annual expiry. Nine months from the last earn or redeem event keeps the balance alive for active shoppers and clears it for lapsed ones, which is exactly the direction you want breakage to run.
• Do not earn on discounted orders above a threshold. Stacking a sale price and a points accrual is how a 55 percent margin becomes a 38 percent one.
• Keep the balance visible at cart, not buried in an account page. An unseen balance produces breakage but no behaviour change, which is the worst of both outcomes because you carry the liability and get none of the retention.

Measuring whether it actually retains

The clean test is a holdout, and most brands refuse to run one because withholding a benefit feels bad. Run it anyway, small and time boxed. Hold 10 percent of new signups out of the programme for one full purchase cycle, matched on acquisition channel and first order value, and compare second order rate and ninety day revenue per customer.

If a holdout is genuinely not possible, use a staggered launch by state or by channel and compare the cohorts before and after. The weakest but still usable method is a regression on member versus non member behaviour with first order value and channel as controls, which at least strips out the most obvious selection bias. What is not acceptable is comparing members to non members raw, because members self select from your best shoppers and the programme will always look like a triumph.

The threshold we use: the programme should lift second order rate by at least three percentage points in the target cohort. Below that, the points are paying for behaviour you already owned, and the same money is better spent on a sharper reorder trigger or a subscription discount that carries a real commitment.

When not to run points at all

Skip the programme if your repurchase cycle is longer than six months, because a balance that sits idle that long has no influence at the moment of decision. Skip it if your catalog is one hero product with no cross sell, because there is nothing for points to unlock. Skip it if your average order value is under Rs 400 and your margin is under 45 percent, because the arithmetic does not leave room. In all three cases a simple, well timed replenishment reminder with a modest fixed benefit will outperform a points ledger and cost far less to build and carry.