Thailand Manufacturing: Forecasting, VMI and Buffers
Long-haul supply from Thailand to the US, Europe or Japan adds four to eight weeks of transit to a manufacturing lead time. That is not a problem in itself — it becomes a problem when planning treats a Thai supplier like a local one. If you are weighing whether to import or manufacture from Thailand, the inventory policy is half the decision.
The real replenishment clock
| Stage | Typical range | Who controls it |
|---|---|---|
| PO to material ready | 1–4 weeks | Factory + tier-2 |
| Production | 2–6 weeks | Factory |
| Booking and port cut-off | 1–2 weeks | Forwarder |
| Ocean transit (Laem Chabang to US West Coast) | 3–4 weeks | Carrier |
| Ocean transit to US East Coast / Europe | 4–6 weeks | Carrier |
| Customs and inland delivery | 1–2 weeks | Broker + 3PL |
Total: commonly 9 to 18 weeks door to door. Air freight compresses this to 3–5 weeks at roughly 6–12x the ocean cost per kg, which is a rescue tool, not a plan.
Sizing safety stock without guessing
Use a simple, defensible formula rather than a gut multiple:
Safety stock = Z × √(lead-time weeks) × weekly demand standard deviation
Where Z is 1.65 for a 95% service level and 2.33 for 99%. Two consequences follow immediately:
- Variability, not average demand, drives the buffer. Stabilising your own promotional calendar is often cheaper than paying for more inventory.
- Buffer grows with the square root of lead time, so shaving two weeks off a 14-week pipeline helps less than most teams assume. Reducing demand noise helps more.
Add a separate, explicit buffer for known step-risks: Songkran (mid-April) and Chinese New Year both slow Thai and regional tier-2 supply, and the monsoon season can disrupt inland trucking.
Three replenishment models that work with Thai suppliers
Rolling forecast plus fixed PO horizon. You share a 6-month rolling forecast monthly; the factory holds long-lead materials for the first 8 weeks; POs are firm inside that window. Simple, works from the first order.
Vendor-managed inventory (VMI) at origin. The factory holds an agreed finished-goods buffer in Thailand against your forecast; you call off containers. Reduces your capital tied up in transit but requires clear ownership, ageing and obsolescence terms.
Bonded or 3PL buffer at destination. Stock lands and sits in a destination warehouse, with duty deferred if bonded. Best service level, highest working capital.
| Model | Working capital | Service level | Best for |
|---|---|---|---|
| Rolling forecast + firm POs | Low | Moderate | Stable B2B demand |
| Origin VMI | Medium | Good | Multi-market call-off |
| Destination buffer | High | Best | Retail and marketplace sellers |
What to write into the agreement
- Forecast cadence, horizon and the accuracy band you commit to.
- Material liability: what the factory may buy on forecast and what you owe if you cancel.
- Buffer size, ownership point, ageing rules and who funds obsolescence.
- Lead-time commitments per product family, with a defined escalation path.
- Capacity reservation for peak months — see capacity planning.
Import versus manufacture, seen through inventory
Buying stock products lets you order smaller and later. Custom manufacture forces you into batch economics: tooling amortisation, minimum material buys and setup costs all reward larger runs. Model the two on total cost including carrying cost, not unit price. A 12% unit-price advantage disappears quickly at 25% annual carrying cost and four extra weeks of pipeline — and equally, it compounds nicely when demand is steady and you order in full containers.
Read next: lead times and production schedules and MOQ strategy.
A 60-day plan to get planning under control
- Map your actual door-to-door lead time from the last five shipments, stage by stage.
- Calculate weekly demand variability per SKU from 12 months of history.
- Set service levels by SKU class; do not use one target for everything.
- Choose a replenishment model per product family and write it into the supply agreement.
- Add calendar buffers for Songkran, Chinese New Year and your own peak.
- Review lead-time performance monthly with the factory using shipment data, not opinions.
Where TUSKO fits
We hold the forecast with the Thai factory, reserve capacity and long-lead materials, police lead-time commitments against real shipment dates, and manage origin buffers where VMI makes sense. You get one point of contact and a planning signal you can trust.
FAQ
How much safety stock do I need for Thai supply?
Calculate it, do not guess: Z × √(lead-time in weeks) × weekly demand standard deviation. For a 14-week pipeline, 95% service and moderate demand noise, buffers of four to eight weeks of average demand are typical.
Is vendor-managed inventory realistic with a Thai factory?
Yes, and it is common for repeat programmes. It needs an agreed buffer level, a defined ownership transfer point, ageing limits and clear obsolescence liability. Without those terms it becomes an argument the first time demand shifts.
How do Thai holidays affect planning?
Songkran in mid-April typically costs several working days, and factories with regional tier-2 suppliers also feel Chinese New Year. Build the loss into your production calendar rather than discovering it in a delayed booking.
Should I use air freight to cover forecast misses?
Only as an exception with a stated approval threshold. Air is roughly 6–12 times ocean cost per kg. If you use it more than a few times a year, your buffer or your forecast process is wrong.
Does a shorter lead time justify domestic manufacturing?
Sometimes, for volatile, bulky or short-shelf-life products. Compare total cost: unit price, freight, duty, carrying cost of the pipeline and the cost of stockouts. See the landed cost model.