Ship-from-Store & Omnichannel Fulfillment: Stores as Supply Nodes | Retail

The conversation that keeps showing up

Rohit: Do you ship from your stores?
Brand Exec: Not yet. We'd like to.
Rohit: What's the blocker?
Brand Exec: Attribution. If digital takes the sale, the store teams disengage. If the store takes it, eCom pushes back. So we keep punting.

I've had versions of that exchange with three brands in the last few weeks, and even inside Mejuri too, before we gradually turned on Ship-from-Store (SFS) across the network.

The pattern is consistent. The blocker is almost never the technology. It's the org. When teams optimize for channel scorecards instead of enterprise outcomes, omnichannel fulfillment dies in attribution debates before it ships a single order. It's rarely anyone's fault: retail, eCommerce, planning, ops, finance, and tech have to operate as one team. If you haven't done that before across those functions, the work can feel too large to start.

It's worth starting anyway. Here's why, and a practical way in.

Stores aren't only demand endpoints. They're supply nodes in a network.

Let inventory move across the network and the whole system gets faster, smarter, and more resilient, and customers get a better experience.

Stores as supply nodes: compounding advantage

Inventory placement is one of the hardest problems in retail: what goes where, and how much? In a channel-siloed model, you have to be precisely right at the node where demand shows up. Too much in a store traps cash and invites markdowns. Too little, and you miss conversion. Online has the same failure mode: the site says "out of stock" while units sit five miles from the customer.

Omnichannel fulfillment changes the math. When a store unit can satisfy digital demand, it stops being stranded. It joins a flexible pool. Soft local demand can be absorbed by digital, and Fulfillment Center (FC) stockout doesn't have to kill the online sale.

Omnichannel fulfillment produces three tangible benefits:

Faster delivery, better customer experienceAt Mejuri, rethinking routing and using our King's Road location as a fulfillment node helped cut UK delivery from roughly 7–9 days to 1–2. Shopify wrote about this work here.
Higher inventory productivityStore inventory becomes multi-channel: sell-through improves, priority transfers and returns to a central FC tend to fall, and cash trapped in the wrong node drops.
Cleaner demand signalsIn silos you measure what each channel could sell given constrained stock, and not what customers wanted. Network fluidity gives planning a truer read by SKU and geography, and forecasts improve over time.

Over time the network is more resilient: imperfect allocation hurts less because inventory can be re-matched to demand. In retail, that forgiveness matters.

Technical problems are solvable. Org problems are where brands stall.

The technical work is real: routing that balances service level, proximity, labor, shipping cost, and inventory health; store inventory accuracy (SFS fails fast if counts aren't trusted); exception paths for labor, packaging, and SLAs. In a modern stack, none of that is novel.

What breaks the model is organizational:

Who gets credit for the order?

If retail and eCommerce are still fighting this, the company is implicitly choosing channel scorekeeping over total performance. Stores resist fulfilling digital orders; digital avoids surfacing store inventory. Leadership gets stuck in attribution instead of improving the network and customer experience.

Why should retail staff pack online orders?

Fair question but with the wrong framing. If the store is part of the fulfillment network, fulfilling demand is customer service. The job isn't "retail plus eCom chores" - it's serving the customer wherever they show up. That reframe has to come from the leadership.

Excessive protective measures surfaces upstream weaknesses

Once you use stores as nodes, buried issues appear: shallow buys, weak localization, inaccurate inventory, slow replenishment, and misaligned KPIs. The natural response is to add more protective measures, but this creates a self-fulfilling prophecy. Technology exposes constraints; the business has to resolve them.

Starter Playbook

You don't turn on SFS across every store on day one. Sequence looks like this:

Pilot 1–3 stores in one real problem marketPick geography where delivery economics or recurrent stockouts are genuinely painful. At Mejuri we expanded UK and Australia before Canada and the US. Use the pilot to learn operations, train staff, build trust, and prove economics on a footprint you can fix when it breaks.
Resist defaulting to per-store safety stock everywhereThe instinct is to protect each door from being "stripped" by digital. At tens or hundreds of locations, per-store buffers recreate the trap omnichannel was meant to solve: units idle while the eComm DC is empty. Prefer no standing stock guardrail; use a temporary guard only while you learn, then remove it. Cleaner signal forces buying and planning to improve - perpetual safety stock often hides their gaps.
Build shared operating visibilityRetail, eCom, planning, ops, and CX need the same views: inventory by node, store fill rates, exception volume, regional delivery times. Different numbers send everyone back to channel silos.
Run one weekly omnichannel reviewSame core group. Review the numbers, ship two concrete changes, repeat. KPI tree should ladder from company-level fulfillment economics down to store service level.

Omnichannel fulfillment isn't a feature toggle. It's a network design: how inventory is placed, how risk is spread, how demand is captured, and how capital is used.

Physical stores are not only demand engines but also a distributed inventory layer closer to customers than most warehouses will ever be. Align the org first; the tech is the easier part.