Which Orders Should You Fulfil Yourself, and Which Should the Platform Handle? (2026)

Every order you route through a quick commerce platform is a customer you already paid to acquire, handed to someone else to serve. Most founders frame this decision as a demand question: where are the shoppers, and how do we get in front of them. That is the wrong axis. The sharper question is narrower and less comfortable. Which of your orders can you actually afford to let someone else own, and which ones quietly fund a competitor’s data moat every time they ship?

The decision is per order, not per channel

Brands treat this as all or nothing. Either you are on quick commerce platforms or you are not. That framing loses money in both directions. A single catalogue holds orders with completely different economics sitting side by side. A ₹180 impulse restock and a ₹4,000 subscription refill do not deserve the same routing logic just because they came from the same brand.

Two variables decide where an order belongs. The first is contribution margin per order: how much is left after product cost, packaging, and fulfilment. The second is how much control the order needs: cold chain risk, high return risk, a regulated SKU, or a customer relationship you want to keep. Route by those two variables and the platform versus own channel question mostly answers itself.

The order routing matrix

Here is the framework we use with brands on our network. Read each order type against margin and control, then route it.

The logic underneath is simple. A quick commerce platform is a reach engine, not a fulfilment strategy. It earns its place when it puts you in front of demand you genuinely cannot reach on your own channel, and when the order is cheap enough that surrendering margin and data does not hurt. The moment an order is profitable, repeatable, or SOP sensitive, the calculation flips towards keeping it in house.

Directional: the profiles above are a routing heuristic, not audited segment data. Validate the margin bands against your own P&L before you act on them.

What the platform quietly takes when it fulfills

When a quick commerce platform owns the order, three things leave the building with it.

Margin goes first. Commissions and fees on quick commerce platforms commonly run from the high teens to well past 30% of order value once you count take rate, fulfilment, and the ad spend needed to stay visible. On a thin margin SKU that is survivable. On your hero product it is your profit.

Directional estimate: platform commission and fee ranges vary widely by platform and category. Confirm against your own contracts before modelling on them.

Data goes second. You receive an anonymized order, not a customer. No name, no repeat purchase signal, no way to bring them back on your terms. Every future order to that same shopper starts from zero. This is the point we made in Marketplace vs Own Channel: What D2C Brands Give Up on Quick Commerce Platforms: the demand looks like yours, but the customer belongs to the platform.

Control goes third, and brands underrate it. Listing on a grocery app turns your packaging and delivery SOPs into suggestions rather than standards. Your cold chain protocol, your unboxing, your damage handling, all of it becomes the platform’s call, executed to the platform’s average and not your standard. For a supplement, a fragile item, or anything where brand consistency is the product, that is a real cost, not a soft one. If you do fulfil on your own channel through a partner, hold them to a hard SLA rather than a rate card. We covered how to read one in How Do You Know If a Fulfilment Partner’s SLA Is Actually Good?.

The break even test

One question settles most routing calls. Does the incremental demand the platform unlocks, demand you could not have captured on your own channel, exceed the margin, data, and control you give up to get it?

If the order would have happened anyway, a repeat buyer or someone who searched your brand by name, the platform is not unlocking demand. It is charging you to serve a customer you already had. Route that order to your own channel. If the order is genuinely new, impulse discovery from a shopper who would never have found you otherwise, and the margin is thin enough to absorb the take rate, the platform earns its cut. Everything in between is a judgement call you now have two clean axes to make.

Where Zippee fits

The routing logic above only works if your own channel can actually deliver at quick commerce speed. That is the gap most brands hit. They know they should keep the profitable orders, but their own website cannot promise 30 minute, 60 minute, or same day delivery, so the platform wins by default.

Zippee closes that gap. We run dark store fulfilment and hyperlocal delivery across 21+ cities including Delhi NCR, Mumbai, Bengaluru, and Hyderabad, giving 100+ D2C brands and marketplaces quick commerce logistics on their own channel. The same speed the platforms offer, delivered on your storefront, with your data, your SOPs, and your brand on the box. That turns the matrix from theory into an operating model: keep the profitable, repeatable, and SOP sensitive orders, deliver them fast, and use platforms only where they genuinely buy you reach.

The takeaway

The routing decision is not about whether quick commerce platforms are good or bad. They are a tool with a specific job: buying reach on orders you can afford to give away. The mistake is handing them the orders that fund your business. Own the profitable ones, deliver them at platform speed, and use the platform only where it genuinely earns its take. Do that, and fulfilment stops being a cost centre and starts being the thing that separates you from every brand that outsourced its customer relationship.

Zippee is the infrastructure that makes it possible: quick commerce speed on your own channel, your data, your brand.

If you’re ready to turn your fulfilment into a competitive advantage, join our waitlist.