How deep should a referral chain go? A margin-first answer
What the numbers say
- 01
A halving payout curve of 8%, 4%, 2%, 1% and 0.5% totals 15.5% of order value at depth 5, against 25% for a flat 5% curve.
Modelled on stated assumptions, summed cumulatively across depths 1 to 5.
- 02
A 14% payout consumes 35.7% of contribution margin on a brand whose contribution margin is 39% of order value.
Modelled on stated assumptions: ₹1,450 AOV, 45% gross margin, ₹83 of order costs giving ₹569 contribution margin.
- 03
The same 14% payout consumes 70% of contribution margin on a brand at 20% contribution margin, which no incrementality rate rescues.
Modelled on stated assumptions, applying the same payout to a thinner margin structure.
- 04
At a 35% ceiling, a brand with 25% contribution margin can afford 8.75% of order value in total payout, which is depth 2 on a halving curve.
Derived from the ceiling rule applied to the modelled margin bands.
- 05
A halving curve converges, so depth 5 costs only 1.5 percentage points more than depth 3, while a flat curve grows without limit.
Derived from the cumulative payout table for the three curve shapes.
Depth gets debated as a marketing question. It is a margin question, and the two produce different answers often enough that the confusion is expensive.
Depth is a margin decision
Payout is charged as a percentage of order value. Affordability is a function of what survives cost of goods, payment fees, shipping subsidy and returns.
Those two numbers diverge by a factor of three between categories, which is why no depth recommendation means anything without a margin attached to it.
On a ₹1,450 order at 45% gross margin, contribution margin after ₹83 of order costs is ₹569, or 39% of order value. A 14% payout is ₹203, which is 35.7% of contribution margin.
Fourteen percent sounds moderate. Thirty-six percent of everything the order earns does not. Both sentences describe the same rupees.
Payout as a percent of order value, depths 1 to 5
Three curve shapes, cumulative.
| Depth | Halving (8/4/2/1/0.5) | Flat 5% | Steep (10/5/2.5/1.25/0.6) |
|---|---|---|---|
| 1 | 8.0% | 5.0% | 10.0% |
| 2 | 12.0% | 10.0% | 15.0% |
| 3 | 14.0% | 15.0% | 17.5% |
| 4 | 15.0% | 20.0% | 18.8% |
| 5 | 15.5% | 25.0% | 19.4% |
The halving curve converges. Each level adds half the previous one, so the total approaches a limit and depth 5 costs only 1.5 points more than depth 3.
The flat curve does not. It grows linearly forever, so every additional level costs as much as the first, and the total is unbounded by design.
The same payouts against contribution margin
The halving curve, expressed against four margin structures.
| Contribution margin as % of AOV | Depth 1 (8%) | Depth 2 (12%) | Depth 3 (14%) |
|---|---|---|---|
| 20% | 40.0% | 60.0% | 70.0% |
| 30% | 26.7% | 40.0% | 46.7% |
| 39% | 20.5% | 30.8% | 35.7% |
| 50% | 16.0% | 24.0% | 28.0% |
The top row is a program that cannot work. Seventy percent of contribution margin going to the chain leaves nothing for the acquisition itself, at any incrementality rate.
The bottom row is comfortable at depth 3 with room to spare.
Identical percentages, opposite verdicts. This table is the entire argument of the article.
The ceiling rule
Keep total chain payout under roughly 35% of contribution margin.
That threshold comes from where the program still clears a plausible breakeven incrementality. Above it, the share of referred orders that must be genuinely new rises past what most programs achieve, and the arithmetic behind that is in the referral math that decides if you make or lose money.
As a formula: maximum payout as a percent of order value equals 0.35 times contribution margin as a percent of order value.
At 39% contribution margin that gives 13.7%, which puts a 14% three-level chain exactly on the line rather than comfortably inside it. That is worth knowing before adding anything.
Depth by margin band
| Contribution margin (% of AOV) | Payout ceiling | Depth on a halving curve | Suggested split |
|---|---|---|---|
| Under 20% | 7.0% | 2 | 5% and 2% |
| 20% to 30% | 7.0% to 10.5% | 2 | 6% and 3% |
| 30% to 38% | 10.5% to 13.3% | 2, or 3 at the top | 7%, 3.5% and 1.75% |
| 38% to 50% | 13.3% to 17.5% | 3 | 8%, 4% and 2% |
| Above 50% | Above 17.5% | 3, with a cap | 8%, 4% and 2% |
Two for most brands. Three for high-margin ones.
Most D2C catalogues land between 25% and 38% contribution margin once order costs are counted honestly, which places the majority in the depth 2 band. Brands that assume they belong in the depth 3 band usually get there by using gross margin instead of contribution margin, which overstates by a modelled ₹83 per order.
SB&R is a Shopify app for chained referral rewards. Every referral link belongs to someone who has already bought. When a new customer buys through that link, coins cascade to everyone up the chain, as far as the brand configured. Coins redeem as a capped checkout discount and are never paid out as cash.
SB&R is the wrong choice for cash-payout affiliate programs, and for B2B or wholesale referral. The ceiling tightens sharply for cash, because cash carries no redemption rate: every rupee issued is a rupee gone, where coins cost only the redeemed share.
Curve shape decides more than depth does
Choose the curve before choosing the depth and the depth question largely settles itself.
A halving curve is self-limiting. Anyone proposing a fourth level next quarter is proposing one percentage point, which is a conversation about rounding rather than about strategy.
A flat curve turns every depth proposal into a genuine budget fight, and those get won by whoever argues best that month. It is how programs drift from 10% to 20% payout without any single decision looking wrong.
The halving shape also tracks what each level contributes: a modelled 100% at level one, roughly 30% at level two, roughly 12% at level three. Paying in proportion to contribution is easy to explain to participants, and a program participants can explain is a program they keep sharing.
The hard cap
Percentages misbehave on outlier orders. A curve that is affordable on a ₹1,450 median order pays ₹1,400 on a ₹10,000 one, and nothing in the percentage logic notices.
Set a maximum rupee payout per order regardless of what the curve computes. Two to three times the payout on your median order is usually defensible.
Anything past depth 3 needs one without exception. Deep chains combined with a wide order value distribution is where payouts appear that nobody modelled and nobody approved.
The second constraint, and when it matters
Affordability is one test. Marginal return is the other: does each additional level bring in more margin than it costs.
That calculation is worked in chained referrals that pay three people, where the marginal return on a halving curve stays positive through depth 4 and turns negative at depth 5 for a brand at the high-margin end.
For most brands the two tests never meet, because the ceiling binds long before marginal return does. A brand at 25% contribution margin reaches its payout ceiling at depth 2 while marginal return would still approve depth 4, and the ceiling wins because it is asking a different question: not whether the level earns its keep, but whether the total is survivable.
Run both. Marginal return tells you when a level stops adding value. The ceiling tells you when the program stops being affordable, and for the majority of catalogues that arrives first, at depth 2, which is a shallower answer than most teams expect and a considerably cheaper one to be wrong about.
Questions people actually ask
What is the right depth for a referral chain?
Two for most brands, three once contribution margin passes roughly 38% of order value, and never deeper without a hard rupee cap per order. Margin decides it rather than order value, because payout is charged against revenue and afforded out of what survives cost of goods, fees, shipping and returns.
Why measure payout against contribution margin instead of revenue?
Because revenue pays for nothing. A 14% payout is comfortable at 39% contribution margin and impossible at 20%, and stating it against revenue hides that difference completely. The same percentage describes two entirely different commitments in two categories.
Should every level pay the same percentage?
No. A halving curve converges, so the total stays bounded however many levels get proposed later, and it matches the modelled contribution each level makes to the sale. Flat curves grow linearly with depth, which turns every future depth discussion into a real budget decision.
Is depth 4 ever worth adding?
Rarely. On a halving curve it adds about one percentage point of payout and a group of participants who cannot explain to a friend why they are being paid. Programs that cannot be described in one sentence stop being shared, and that cost never appears in the model.
What is a hard cap and when does it matter?
A maximum rupee payout per order regardless of what the percentages compute to. It matters most where order values vary widely, because a curve that is affordable on a median order can produce an alarming payout on an outlier. Anything past depth 3 should carry one.