How to build an inventory management system: barcodes, multi-warehouse, and e-commerce/POS integration (cost + 90-day roadmap)

Opening pain point + real scene

A clothing brand with two stores plus an online shop has had mismatched inventory for half a year: the store POS shows in-stock, the e-commerce backend shows out-of-stock, so the same jacket sells online and offline at once and someone gets an apology and refund. The owner spends 6–8 hours a week manually reconciling three Excel files. This isn''t "should we digitize" — it''s "every day without integration is a day of leaking money." An inventory management system solves exactly this.

Suitable vs unsuitable

Suitable:

2+ sales channels (store + e-commerce + LINE) needing inventory sync
200+ SKUs where Excel is starting to break
Multiple warehouses/stores needing transfers and per-location stock
Need real-time cost, margin, and safety-stock reports

Unsuitable (don''t build yet):

Under 50 SKUs, single channel — Excel or a POS''s built-in inventory is enough
Pure services, no physical stock
Just wanting to "look professional" with no real reconciliation pain

Alternatives matrix

Option	Pros	Cons	Cost band
Packaged inventory software	Full features, has support	Low customization; integrating your own e-commerce often costs extra	Monthly NT$2,000–8,000 + onboarding
E-commerce built-in inventory (Shopify/WooCommerce)	Native to the store, fast start	Weak multi-store/warehouse; cross-platform sync hard	Included in platform fee
Excel/Google Sheets DIY	Zero cost, flexible	Error-prone, no real-time sync, high manual reconciliation	NT$0 + heavy labor
Custom system (ScriptWalker)	Fits your channels/flow, integrates existing e-commerce/POS	Higher upfront; requirements must be clear	From NT$120,000

Full process breakdown (tools + deliverables)

Stage 1: Requirements & audit (1–2 wks) — audit SKUs, channels, warehouses, existing systems. Deliverable: inventory flow diagram (Figma/Notion), SKU naming rules.
Stage 2: Data model & barcode planning (1 wk) — design product, batch, warehouse, stock-movement tables; plan barcodes per the GS1 barcode standard. Deliverable: DB schema, barcode spec.
Stage 3: Core build (3–5 wks) — purchasing, sales, transfers, stocktakes, safety-stock alerts; backend in Laravel. Deliverable: working admin.
Stage 4: Channel integration (2–3 wks) — connect e-commerce (Shopify/WooCommerce API), POS, payment reconciliation. Deliverable: two-way inventory sync.
Stage 5: Testing & training (1–2 wks) — pilot with real data, write SOPs, train store staff. Deliverable: acceptance checklist, manual.

Full cost breakdown

System build: from NT$120,000 (by channel count and customization)
Payment reconciliation: with ECPay, ~2.8% payment fee; integration hours extra
Barcode scanner/label printer (hardware): NT$3,000–15,000
Cloud hosting: monthly NT$1,500–4,000
Hidden costs: e-commerce API transaction cuts, third-party logistics API, and data cleansing (old Excel import) are often underestimated — together another NT$20,000–40,000
Post-launch maintenance: recommended from NT$8,000/month

Reality vs client imagination

Client thinks: just import the Excel. Reality: old data has messy SKU naming and multiple spellings per item; cleansing often eats 20% of project hours.
Client thinks: sync is instant. Reality: most e-commerce APIs are rate-limited, so sync is "near real-time" (seconds to minutes); high concurrency needs a queue.
Client thinks: it''s accurate once live. Reality: book-vs-physical gaps come from human operation; the first 1–2 months need stocktakes to stabilize.

Common traps and how to avoid them

Inconsistent SKU naming → set naming rules before starting and clean once.
No safety-stock alerts → build reorder thresholds and notifications in from the start.
One-way e-commerce sync only → make it two-way (sell deducts, restock adds) or you''ll still oversell.
No transfer flow → multi-store must have transfer orders or locations never reconcile.
Ignoring return restocking → if returns don''t reverse stock, the books drift.

Success metrics + 90-day roadmap

Day 30: inventory accuracy (book vs count) 95%+, oversell events at zero.
Day 60: weekly reconciliation drops from 6–8 hours to under 1; restock via alerts, not manual checks.
Day 90: per-SKU margin and turnover reports; start ordering by data.

Decision checklist

☐ I have 2+ channels needing inventory sync
☐ My SKUs are numerous enough that Excel breaks
☐ I have multiple stores/warehouses needing transfers
☐ I spend 3+ hours/week reconciling manually
☐ I''ve had oversells or stock misjudgments
☐ I want real-time margin and turnover reports
☐ I''m willing to spend time cleansing old data
☐ I can assign an internal point person
☐ I understand sync is near real-time, not absolute
☐ I''m ready to correct via stocktakes the first 2 months

Check more than 6 and a custom inventory system is worth serious evaluation.

FAQ

Can I just connect my existing e-commerce without switching platforms?

Yes. Usually we add an inventory hub on top of your existing Shopify/WooCommerce, syncing two-way via API — no migration needed.

How long does rollout take?

Single store + single e-commerce ~6–8 weeks; multi-store/warehouse/channel ~10–14 weeks. The time sink is usually data cleansing and training, not development.

Can I start with a minimum version and expand?

Recommended. Build single source of truth + two-way sync + oversell prevention first; put reports and multi-warehouse transfers in phase two for faster payback.

Who''s responsible if it breaks?

We offer post-launch maintenance plans with monitoring, backups, and troubleshooting, and deliver source code and docs to avoid vendor lock-in.

Call to action

Want to know if your channel mix fits a custom inventory system? Book a free 30-minute consult — we''ll sketch an inventory flow diagram and a rough quote: