A brand decides to sell direct. The reasoning is sound: own the customer, see the demand, launch a product without waiting for anyone to stock it.
Then the first order arrives from a town where a distributor has spent fifteen years building coverage. The brand ships it from a central warehouse, three days late, at a price the distributor cannot match. The distributor notices. The next quarter's primary order is smaller, and nobody writes down why.
This is the part of D2C that gets discussed as a "channel conflict problem", as though it were a communications exercise. It is not. It is a routing decision that most commerce software cannot express.
The question is who fulfils, not who sells
Strip the emotion out and the disagreement is narrow. The brand wants the storefront, the customer relationship and the demand signal. The distributor wants the fulfilment and the margin in the territory they built.
Those are not the same thing, and they do not have to go to the same party. A consumer order can be raised on the brand's storefront, priced by the brand's rules — and then fulfilled by whoever actually serves that pincode, with the margin following the fulfilment.
The reason this is rare in software is that it requires the order to carry a selling organization that is not the tenant running the storefront. Most systems assume those are the same, and the assumption is invisible until you try to do this.
Serviceability is declared, not inferred
The routing has to be based on something concrete. "Nearest warehouse" is a bad proxy — a distributor forty kilometres away with a delivery van beats one twenty kilometres away without one.
So each partner declares what they will actually deliver to: a set of pincodes, or a drawn polygon where a pincode is too coarse to be useful. Overlaps are resolved by an explicit priority rather than an implicit rule, which matters because overlaps are common and the tie-break will eventually be questioned.
The immediate benefit is on the storefront: a consumer types their pincode and learns straight away whether anyone serves them, instead of discovering it after payment. The larger benefit is that the brand now has a map of demand it cannot serve — which pincodes are asking, and for what. That is a coverage plan, generated by the storefront rather than by a survey.
Commission has to be defensible, which means historical
The arrangement lives or dies on settlement. Every partner-fulfilled programme we have seen fail did so in a spreadsheet, monthly, with someone reconstructing commission from exports and then defending the number in a meeting.
Two properties make it defensible instead.
It accrues as it happens. Each order produces a settlement line at the time of sale, rolling into a period statement with gross, commission, net and order count. Nothing is reconstructed later, because nothing needs to be.
The rate is captured, not looked up. Each line carries the commission percentage in force when that order was sold. Renegotiating a partner's rate in March does not silently restate February. This sounds like a small implementation detail and is in fact the entire basis of trust in the arrangement: a statement that can change retroactively is a statement nobody believes.
What the partner actually experiences
The other common failure is giving partners a separate portal to check. It gets checked for a fortnight and then it does not.
An order routed to a partner should land where their orders already land — the same list, the same fulfilment tooling, the same mobile app they use for everything else. If a partner has to remember to look somewhere new, the programme depends on a habit, and habits decay.
Where this leaves the brand
With less than they imagined and more than they had. Not every consumer order, and not every rupee of margin — but the storefront, the customer relationship, the pricing policy, the demand map, and a channel that is not quietly working against the direct programme.
The brands that sustain a direct channel are generally not the ones that beat their distributors to the customer. They are the ones that stopped treating the last mile as something to disintermediate and started treating it as something to route.