A single working surface for the Shenzhen-to-Bangkok trade on 8/16 GB DDR4 SODIMM at 2666/3200 MT/s. Landed cost live-calculated with customs scenarios, freight and FX; supplier scorecard, staged batch progression, and daily news, in one page.
Downside before upside. Cash at risk, three scenarios (Base / Stress / Kill), and stop-rule status — one screen that answers "Buy / Sample / Do not import." All numbers derived from the Assumptions Ledger below; edit an assumption, every tile here recomputes.
FX pulled live from open.er-api.com. Landed cost is for the star SKU (16 GB 3200) at the workbook's used-refurb wholesale of $22 with Form E and sea LCL freight.
Change any input; the breakdown recomputes. Post-Jan-2026 Thai customs: no more THB 1,500 exemption — 7% VAT applies on CIF+duty on every parcel.
A Stage 1–2 sample or pilot is light and small (typically < 5 kg). At that size LCL sea's fixed origin/destination fees eat the per-stick advantage; air express usually wins on total cost + speed. All figures shown are indicative bands — replace with written forwarder quotes before booking.
| Mode | Carrier examples | Transit | Per-stick freight | Sample/pilot batch cost | Notes |
|---|---|---|---|---|---|
| Air express Recommended (Stage 1–2) | DHL, FedEx, UPS, SF Express | 3–5 days | $2.15–$2.85 | $150–$200 | Door-to-door, tracked, Form E friendly, no consolidation delay |
| Sea LCL | Freight forwarder (Superbuy, Wise-Cargo, Bansar) | 15–25 days | $1.50 (at ≥50 pcs) | $180–$260* | *Fixed origin/destination fees inflate small parcels; economical from ~500 pcs |
| China–Thai overland | Kerry, Best, J&T cross-border | 5–10 days | $2.00–$4.00/kg | $120–$180 | Cheapest for small parcels; documentation is spottier — verify Form E acceptance |
All-in freight figures include origin pickup, brokerage, and Thai last-mile. Confirm with 2–3 forwarder quotes before booking; rates move week-to-week.
Three price tracks per SKU: original workbook (Jul 2026), current Sept 2026 wholesale range from Alibaba/refurb dealer scans, and the new landed cost implied at midpoint.
DDR4 SODIMM kit contract prices are up 277–380% since Q1 2025 (TrendForce). Q1 2026 conventional DRAM +90–95% QoQ; Q2 forecast +58–63%; Q3 DDR4 8Gb +50%. Spot is trading 172% above contract — sign of severe scarcity, not price discovery.
Samsung + SK Hynix delayed DDR4 phase-out into 2026 (some further) — they're "milking the shortage" (wccftech, Sep 2025 → confirmed Jan 2026). Micron transferred some DDR4 to Fab 6 Virginia without adding capacity.
What this means for you (market note, not directive): your July workbook wholesale numbers ($11–52 range) are 40–80% below current observed listings. Whether Thai retail has moved in lockstep requires SKU-level verification from live Thai marketplace listings — the 5.2× multiplier used below is modeled, not verified. Do not treat the retail figures as investment-grade until you populate the Benchmarks tab with dated observed listings.
MODELEDTH retail figures below are 5.2× wholesale midpoints, not observed Thai listings. Do not use for pricing decisions. Verify by pulling 10+ Shopee/Lazada listings per SKU, then compute P25 / median / P75 per condition (new/used/refurb) and per format (single/kit). Until then, treat "Margin at midpt" as an upper bound, not a target.
| SKU (DDR4 SODIMM) | Workbook Jul '26 (USD) | Sept '26 wholesale (USD) | Sept '26 used-refurb (USD) | TH retail MODELED | Margin upper bound | Priority |
|---|---|---|---|---|---|---|
| 8 GB · 2666 | $11 | $28 – $42 | $18 – $28 | 1,300 – 1,700 | ~24% | High ★ |
| 8 GB · 3200 | $13 | $37 – $45 | $22 – $32 | 1,500 – 1,950 | ~26% | High ★ |
| 16 GB · 2666 | $19 | $58 – $80 | $34 – $50 | 2,400 – 3,100 | ~30% | High ★★ |
| 16 GB · 3200 ★ star SKU | $22 | $64 – $95 | $40 – $58 | 2,800 – 3,700 | ~32% | BEST ★★★ |
| 32 GB · 3200 | $52 | $135 – $195 | $82 – $120 | 5,200 – 6,800 | ~25% | Grow ★★ |
| DDR5 SODIMM · 16 GB · 5600 (surging) | — | $175 – $240 | — | 6,500 – 8,400 | ~22% | Phase 1.5 ★★ |
| DDR5 SODIMM · 32 GB · 5600 (watch — TechRadar flagged $500 module scenario) | — | $310 – $460 | — | 11,500 – 15,500 | ~20% | Watch |
Data as of Sept 10, 2026 · WebSearch-refreshed. Wholesale ranges: Alibaba MOQ-dependent bands (lower = ≥500 pcs, higher = ≤50 pcs, brand-name IC premium). Used-refurb: bigdatasupplyinc / KimoStore / Jawa / eBay lot pricing. Thai retail: modelled at 5.2× Sept wholesale midpoint (matches your workbook's 3.7× multiplier on the older $22 → 2,300 THB, escalated by the same 40% Thai retail move). Verify with 3 live supplier RFQs before committing capital.
Sources: TrendForce spot Jun 2026 · wccftech Q3 2026 +50% · TechPowerUp server DRAM +50% · Tom's Hardware DDR4 lifespan · TechRadar $500 32GB DDR5 · wccftech phase-out delay · Alibaba wholesale scan
Every platform bites twice: commission on the sale, and ad spend to win the click. Set your monthly units and channel mix; the table computes each channel's take, ad burn, and the blended net you keep.
| Channel | Mix % | Units | Revenue (THB) | Commission % | Txn fee % | Service/Xtra % | Ad spend % | Per-order fixed (THB) | Total take (THB) | Take % |
|---|---|---|---|---|---|---|---|---|---|---|
| Blended · rate % | — | — | — | — | — | — | — | — | — | — |
| Column subtotal (THB) | expenses by category ↓ | — | — | — | — | — | — | ← grand total | ||
Gross revenue minus commissions, txn fees, service, ads and per-order fixed. This is what hits your account before cost-of-goods.
Uses landed COGS at current calculator settings. If this dips below ~15%, either raise retail or shift mix toward lower-fee channels (Kaidee, FB).
Total ad spend ÷ monthly units. Watch it against your gross-profit-per-unit — if CPA > GP/unit, you're paying to lose money.
How to use: edit any cell in the table (mix %, fee rates, ads) — everything recomputes instantly. Mix % must sum to 100 (indicator on total row). Defaults are current Thai e-commerce benchmarks; verify at each platform's current fee page since Shopee/Lazada retune quarterly.
Fee references: Shopee TH seller fees · Lazada TH seller centre · TikTok Shop TH · FB Marketplace · Kaidee
Every input that drives a number on this page, with its confidence tag. Green badges = verified live. Yellow = estimated. Red = missing / stale. Edit values freely; the Dashboard, Scenarios, and Summary all pull from here.
| Variable | Value | Unit | Source | Confidence | Note |
|---|
The ledger stores your source-of-truth. If a value here is MISSING or STALE, the Summary tab flags the batch as unsafe to place.
Two-part landed model: 3 customs scenarios (Form E accepted / rejected / delayed) × yield-based per-saleable-unit costing (accounts for DOA and warranty reserve, not just ordered units). This replaces the naïve "landed / qty" from the top-of-page calculator.
0% duty · 7% VAT on CIF · standard broker
5% duty · 7% VAT on CIF+duty · same broker
5% duty + demurrage + extra brokerage + 15-day lag
Reality: not every stick you order becomes a sellable one. Intake QC rejects some, burn-in fails others, warranty reserves take a slice. This card computes the honest cost per saleable unit.
Three side-by-side scenarios — Base (median assumptions, Form E works), Stress (Form E rejected, higher DOA, price 12% under median), Kill (customs surprise, 90+ day sell-through, margin below floor). Each ends with a Buy / Sample / Skip verdict. Stop-rule gates below tell you when to halt regardless.
These are the six hard gates. Any failure = do-not-scale. Sticky on desktop; scannable on mobile.
Left panel pre-loaded with Feifan's July 2026 OEM sheet (all DDR4 SKUs, USD + RMB as quoted). Enter a competing supplier's quote on the right — everything shown in USD, RMB & THB, cheaper cells highlighted green, winner named below.
Enter competing supplier prices on the right to see winner.
Comparison uses landed-unit-cost equivalence in THB (multiplied by qty). FX conversions use your current snapshot rates. Change qty on either side to see how order-mix affects the winner — sometimes supplier A wins on 8GB but supplier B wins on 16GB, and the mix decides.
Left: Feifan's July 2026 OEM sheet — all DDR2/DDR3/DDR4/DDR5 SKUs, exact prices locked as reference, plus live THB conversion. Right: same rows starting from the same prices, but every cell editable — model a negotiated rate, a competing quote, or a stress test. Grand-total delta shown below.
Enter quantities on either side to see the running total.
Delta = Scenario grand total − Reference grand total. Negative delta means your scenario is cheaper than the Feifan reference (you negotiated a saving). Positive delta means costlier. Both sides use the current FX shown above.
Reframed after review: the 70-stick order is a scale batch, not a tryout. Real risk-controlled entry uses four stages, each with an explicit gate to proceed. Start at Stage 1 (~$150–$450), not Stage 4.
| Stage | Qty | Capital (THB) | Objective | Gate to proceed |
|---|---|---|---|---|
| 1 · Supplier sample START HERE | 6–10 | ~5,000–15,000 | Validate supplier quality, docs, SPD/spec match | ≥ 95% QC yield · correct spec · Form E draft received · watermarked photos |
| 2 · Sales pilot | 12–20 | ~20,000–35,000 | Validate price, listing conversion, warranty support, returns | ≥ 70% sold in 30 days at target contribution ≥ THB 300/stick |
| 3 · Controlled replenishment | 25–40 | ~40,000–65,000 | Confirm repeatability, real returns rate, warranty burn | Returns/warranty within reserve · no customs surprise · Form E accepted at Thai border |
| 4 · Scale batch SHOWN BELOW | 50+ | ~100,000+ | Improve sourcing economics via bulk pricing | Two successful cycles at Stage 3 with recorded results |
Where you sit today: if this is your first import cycle, you are at Stage 1. The composition table below is Stage 4 planning — do not use it as your first PO. Capital exposure at Stage 1 is ~$150–$450, not ~$3,190.
Open Alibaba Trade Assurance. Send RFQ (below) to 5–8 verified suppliers. Order 3–5 pcs per top-2 SKUs. MemTest86 4-pass + TM5 1 hr on every stick. Photograph + serial log. Get Form E draft from supplier.
Advance only if Stage 1 QC ≥ 95% and Form E draft is real. List tested pilot batch with 90-day warranty. Validate listing conversion, ad spend efficacy, customer questions, returns process.
Repeat with same supplier + same SKUs. Track actual returns rate against your 5% reserve. Confirm Form E was accepted by Thai customs (not just issued) on the first import.
Only after two successful Stage 3 cycles. This is when the composition table below applies. Multi-channel launch (Shopee, Lazada, FB, Kaidee, TikTok).
| SKU | Qty | Landed (THB) | Total |
|---|---|---|---|
| 8 GB 2666 | 10 | 1,026 | 10,262 |
| 8 GB 3200 | 30 | 1,191 | 35,732 |
| 16 GB 2666 | 15 | 1,819 | 27,281 |
| 16 GB 3200 ★ | 15 | 2,111 | 31,672 |
| Total | 70 | — | ~104,947 |
Expected revenue ~156,750 THB · gross profit ~31,300 THB after 13.09% platform fee · ~30% ROI on capital before fixed costs · ~447 THB profit/stick. Capital ~$3,190 USD at current FX — up from ~$1,620 in the July workbook because global wholesale moved.
Subject: RFQ — DDR4 SODIMM 8/16GB 2666/3200 to Thailand (Stage 1 sample + staged replenishment) Hello, We are a Bangkok-based retailer sourcing DDR4 SODIMM for the Thai e-commerce market. This RFQ is for a Stage 1 supplier-qualification sample. Successful sample QC opens a documented replenishment path (Stage 2 pilot 12–20 pcs, Stage 3 controlled 25–40 pcs, Stage 4 scale 50+ pcs). SKUs (all DDR4 SODIMM, 260-pin, 1.2V): 1) 8 GB · 2666 MT/s — qty 3 sample / 50 pcs Stage 3 / 100 pcs Stage 4 2) 8 GB · 3200 MT/s — qty 3 sample / 50 pcs Stage 3 / 200 pcs Stage 4 3) 16 GB · 2666 MT/s — qty 2 sample / 30 pcs Stage 3 / 100 pcs Stage 4 4) 16 GB · 3200 MT/s — qty 2 sample / 30 pcs Stage 3 / 100 pcs Stage 4 Please confirm for each SKU: • Unit price EXW Shenzhen (USD) at each stage quantity • IC brand (SK Hynix / Micron / Samsung required; no unbranded) • Condition: new / used-tested / refurbished — state clearly, no ambiguous "OEM" • Testing method: MemTest86 pass count, burn-in hours, error tolerance • Warranty: minimum 90-day DOA replacement with return shipping paid • Packaging: individually anti-static, retail box optional Trade terms: • Payment: Alibaba Trade Assurance, 30% deposit / 70% on BL or air waybill • Form E certificate (HS 8473.30.30) MUST be provided; if you cannot issue Form E, quote must include a written declaration and price will be adjusted for 5% duty • Commercial invoice and packing list at true declared value Logistics: quote both DHL/FedEx air express and sea LCL to Bangkok (BKK). Stage-2 pilot order is placed only if sample QC yields ≥ 95% (reject rate ≤ 5%). Escalation to Stage 3/4 follows documented gates on returns rate, Form E acceptance at Thai customs, and sell-through in Thai marketplaces. Please reply with: 1. Filled quotation table (per-SKU, per-stage pricing) 2. Real photos of current stock (watermarked, dated — no stock images) 3. Sample MemTest86 report for one stick of each SKU 4. Business licence and Thailand export history Best regards, [Your name] [Company] · Bangkok, Thailand
Answer 10 questions per supplier. The score updates live. 75+ = GO for Stage 1 sample · 55–74 = smaller Stage 1 sample only · <55 = SKIP.
Honest read on the arbitrage window. The shortage is real through 2027 but not eternal — DDR4 inventory should not roll into 2028.
Computed from stop-rules pass rate × verified-assumption ratio × stress-case survival. Changes when assumptions change.
Cumulative units to cover 8,000 THB/mo fixed cost at current mix. Recomputes with calculator inputs.
Net profit on ~1.21M THB capital deployed and ~1,499 units sold. Conservative vs the original PDF's 3M THB projection.
Single-page executive read: verdict, headline metrics, the conditions the verdict depends on, and next actions. Every number here traces back to the Assumptions Ledger.
Fill the Assumptions Ledger and set your batch quantities to render the verdict.
Verdict depends on all six stop-rules passing.
Total THB tied up from deposit through inventory receipt.
After landed cost, platform fees, ads, packaging, DOA reserve.
Days to sell full batch at target sell-through.
Kill-case exposure if customs, QC, and market all miss.
Ratio of verified-source variables to total.
DRAM/HBM/tariff headlines refresh on page load via Google News RSS. Equities link out to Yahoo Finance for the three suppliers whose capacity decisions drive spot prices.