Replacing ShipStation with a custom multi-channel order platform

The problem

We were running five storefronts — Amazon, Walmart Marketplace, Etsy, Shopify, and TikTok Shop — through ShipStation plus a patchwork of spreadsheets and manual CSV exports. On paper it worked. In practice it cost us money and hours every single day.

Orders landed in five different dashboards with five different data shapes. ShipStation's per-order pricing and add-on fees kept climbing as volume grew — and at roughly 5,000 orders a month, per-order pricing is not a rounding error. Inventory drift between channels meant oversells and cancellations, which on Amazon and Walmart directly threaten your seller metrics.

The approach

I'd been automating pieces of this for a while. The step change was deciding to replace the hub itself. Working with one other developer, I helped build a single order platform that ingests every channel through its native API, normalizes orders into one schema, and drives fulfillment from a unified queue.

The design goal was boring on purpose: one screen, one source of truth, zero CSV exports. Everything else followed from that.

Sales channels

• Amazon API
• Walmart Marketplace API
• Etsy API
• Shopify API
• TikTok Shop API

↓↔

Multi-channel Order Platform

One normalized queue · PHP

↓↔

Fulfillment

• Unified order queue
• Label generation
• Shipment tracking
• Inventory sync

Key decisions:

• Native APIs over a middleware tax. Pulling directly from each channel's API removed both the per-order fee and the sync lag that caused oversells.
• Built to fit the operation, not the average customer. Off-the-shelf tools are built for the median seller. Ours is built for our catalog, our channel mix, and our shipping profile — which is why it's both cheaper and better at the specific job.
• One inventory picture. A shared availability number pushes back out to the channels, so selling the last unit in one place decrements it everywhere.

Before

Before — 5 dashboards, CSV exports

After

After — one unified order queue

The result

The platform replaced ShipStation and the manual work around it outright.

The number I care about most isn't the $24k — it's that oversells effectively went to zero, which kept our Amazon and Walmart metrics clean.

• ~$24,000 per year saved in subscription, per-order, and add-on fees.
• Oversells effectively eliminated via the shared inventory picture, protecting seller metrics across marketplaces.
• 99% on-time shipping sustained at roughly 5,000 orders a month.
• One queue instead of five dashboards — batch label generation replaced tab-hopping and manual exports.

It's the clearest example of how I work: when a tool stops fitting the operation, I'm comfortable helping build the one that does.

Stack