How to Build a Simple Order Fulfillment Process for E-commerce

oms

Most stores don’t have a problem with “order fulfillment” as such — they have a problem with doing it slightly differently every time. One day a parcel ships within an hour, the next it sits half a day because someone forgot to check the payment. Sometimes the tracking number reaches the customer, sometimes it doesn’t. The result is chaos that doesn’t come from a lack of work, but from a lack of process. And a process is nothing more than a fixed, repeatable sequence of steps that everyone in the company performs the same way — including when you’re not there.

This article is a concrete implementation guide. We’ll walk through the five stages of an order (intake → verification → picking → shipping → after-sales), who owns what, where to plug in automation and which metrics to watch so you know whether the process actually works. No theory — written so you can roll it out this week.

Before you design the process: map what you do today

Don’t start from an “ideal” process out of a textbook. Start by writing down what happens right now — from the moment an order comes in to the moment the customer receives it. Take a sheet of paper or a document and list every step, every person and every place someone has to check (store panel, email, warehouse, courier system). The list usually turns out longer than you’d expect — and two things immediately stand out:

bottlenecks — steps where the order “waits” (most often payment verification),
context switching — how many different windows and logins it takes to handle one order.

Only with this map in front of you can you decide deliberately what to simplify, what to automate and what to leave to a human. The process is built on reality, not on a hunch.

The five stages every order goes through

Regardless of industry, every order passes through the same skeleton. Your exceptions (pre-orders, cash on delivery, personalization) attach to it as branches — but the core stays the same.

Intake — the order lands in your system. The key question: can you see it in one place, regardless of channel? If you have to check three panels, the process starts in chaos.
Verification — confirming payment, address correctness, stock availability and the absence of warning signals (suspicious buyer, unusual quantity). This is the quality gate of the whole process.
Picking — gathering the goods from the warehouse, checking them against the order, packing. This is where most physical errors happen: wrong item, missing line, damage.
Shipping — dispatching the parcel, generating the label, saving the tracking number and passing it to the customer. The moment the order “leaves the company.”
After-sales — confirming delivery, requesting a review, handling returns and complaints. The stage most often skipped, yet decisive for whether the customer comes back.

Each arrow between stages is a transition that can be described with a rule and — if it’s deterministic — handed to automation. But first it has to be clear who owns each stage.

Roles: who owns what

A process without assigned roles is a wish list. Even in a one-person business it’s worth naming the roles — because when you hire your first person, you’ll know exactly what to hand over. The minimum split:

Order handling — watches the intake and verification stages, deals with exceptions and customer contact. This role decides whether an order “moves on” or goes to review.
Warehouse / packing — responsible for picking and physically preparing the parcel. Its metric is the parcel’s match to the order, not speed at any cost.
Shipping — dispatch, labels, passing on the tracking number. In a small store this merges with the warehouse.
After-sales — returns, complaints, reviews. Requires decision-making permissions (refund, exchange).

It matters that roles have matching permissions in the system. A warehouse person doesn’t need to see billing data, and a packer shouldn’t be able to delete orders. Sellaro supports multiple users and roles, and each customer’s data is isolated in a separate database schema (schema-per-tenant) — so dividing the work doesn’t mean handing everyone access to everything.

Where to plug in automation — and where to keep a human

The rule is simple: automation handles the rule, a human handles the exception. Don’t automate everything at once — plug automation into the transitions that are fully deterministic and repeatable:

after payment is confirmed → change status to “in fulfillment” + confirmation email,
after the parcel is dispatched → status “shipped” + save the tracking number + SMS with a tracking link,
after delivery (if the channel reports it) → an automatic review request a few days later,
on missing stock or a wrong address → flag “needs review” and alert the team (don’t push it forward).

Sellaro’s automation engine runs on domain events in a simple WHEN → IF → THEN formula: when something happens, if a condition holds, then perform an action. The available actions today are email (SMTP) and SMS notifications, webhooks (HMAC-signed with retries) and a log entry. Which specific status transitions pay off fastest, we broke down in a separate piece on order status automation.

What not to hand to automation unsupervised: returns and complaints, “needs review” orders, cash-on-delivery and pre-orders. A well-designed process catches and holds these cases so a human deals only with the exceptions — instead of clicking through every single order.

A good process isn’t about everything happening by itself. It’s about the routine happening by itself while your attention goes where it’s genuinely needed — to the exceptions.

Metrics: how you know the process works

A process without metrics is an impression, not a fact. You don’t need elaborate analytics — a handful of numbers you track weekly is enough:

Fulfillment time — from intake to dispatch. The single most important indicator.
Share of “needs review” orders — how many fall out of the automated path. Rising = something’s breaking.
Number of picking errors — wrong/missing lines per parcel. A measure of warehouse quality.
Return and complaint rate — a signal of overall process quality, not just the product.
After-sales response time — how fast a customer gets an answer on a return or question.

A single, normalized view of orders across all channels helps a lot here — because you’re measuring one process, not five separate panels. On how to systematically reduce mistakes at each stage, we wrote more in the guide on how to reduce order processing errors.

Frequently asked questions

Where do I start building an order fulfillment process?

By mapping what you do today — writing down every step from an order coming in to the customer receiving it. Only on that basis will you see the bottlenecks and places to simplify. Then name the five stages (intake, verification, picking, shipping, after-sales) and assign roles to them.

Is it worth designing a process if I run the store solo?

Yes. A process written for one person is a ready-made onboarding manual for your first employee and a safeguard for when you’re away. Naming roles “on paper” doesn’t require hiring anyone — and when the moment comes, handing over duties takes hours, not weeks.

Which stages should I automate first?

The fully deterministic transitions: the status change after payment is confirmed and after the parcel is dispatched, together with sending the tracking number to the customer. That’s the cheapest start with the highest return. Leave returns, complaints and “needs review” orders to a human — automation should only flag them.

Does Sellaro generate labels and dispatch parcels for me?

Not today. Sellaro centralizes orders, maintains their status on its side and runs automations (email, SMS, webhook, log). Courier label generation and integrations with InPost and other carriers are on the roadmap and we add them on request. Store integrations are read-only — Sellaro reads the data but does not write changes back to the store.

Summary

A simple order fulfillment process is five repeatable stages — intake, verification, picking, shipping, after-sales — with clearly assigned roles, automation plugged into the deterministic steps and a few metrics you track weekly. Start by mapping reality, hand the routine to automation, keep the exceptions for yourself, and measure fulfillment time and the share of “needs review” orders. The rest is consistency and discipline.

Sellaro provides today a central, normalized view of orders across all connected channels (PrestaShop, Sylius, WooCommerce — ready; Allegro, Amazon, Shopify — on the roadmap), an event-based automation engine and multi-user, role-based access — the foundation on which such a process can actually be sustained. See Sellaro pricing — a flat fee with a generous limit, all integrations included in the plan and 0% commission on sales value. Missing an integration you need? Let us know — we’ll add it for free.