Technicals — Green Bean Sourcing & Timeline
Info
KNNO sources Indonesian-first in year 1. We go deep on a few producers rather than wide across many. Quality > variety > price, in that order.
1. Sourcing Principles (the immovable rules)
- Indonesia first. Year 1 — at minimum 90 % of green volume from Indonesian origins.
- Producer-named lots only. No anonymous “Sumatra Mandheling” commodity grade.
- Process transparency. We must know how it was fermented, for how long, and (if applicable) what was added.
- Sample before commit. No buying a 30 kg lot without 250–500 g sample first.
- Roast date matters more than purchase price. Better to pay slightly more for a fresher harvest than save 10 % on stale green.
- Build a 3-region portfolio. Aceh + West Java + Bali (or Sulawesi) — diversification against harvest failure.
- Prioritise small/micro producers. They’re more open to direct relationships and small volumes.
2. The Indonesian Co-Ferment Producer Landscape
Tier S — Established producers with international reputation
| Producer / Estate | Region | Speciality | How to access |
|---|---|---|---|
| Frinsa Estate (Wildan Mustofa) | West Java | Anaerobic naturals, Frinsa Manis (strawberry/jasmine notes) | Direct via estate (Bahasa email); also via Indonesian exporters; international roasters resell |
| Klasik Beans Cooperative | West Java | Yeast-inoculated, anaerobic, micro-lots | Direct via cooperative office in Bandung |
| Karana Coffee | Java/Bali | Carbonic maceration, anaerobic | Direct producer outreach |
| Java Halu | West Java | Volcanic-soil washed + anaerobic | Through Indonesian exporters |
| Aceh Gayo cooperatives (Permata Gayo, Ketiara, Arinagata) | Aceh | Washed, wet-hulled, increasingly anaerobic | Cooperative direct or via Toba Coffee/Sumatra Coffee |
Tier A — Smaller producers / micro-lots (build relationships year 1)
- Burni Telong (Aceh) — anaerobic experiments
- Wahana Estate (Sumatra) — extensive co-ferment program
- Toarco / Toraja Sapan (Sulawesi) — washed + experimental
- Pwani / Bajawa producers (Flores) — emerging anaerobic
- Kintamani micro-producers (Bali) — CM, wine yeast experiments
Tier B — Indonesian exporters (good for samples, less for relationships)
- Indo Sumatra Coffee Trading
- PT Sumatra Coffee (sumatracoffee.co.id)
- FNB Coffee (fnb.coffee) — broad menu, some co-ferments
- Indonesia Specialty Coffee (specialtycoffee.id) — content-heavy + selling
International importers with Indonesian co-ferment lots (backup channel)
| Importer (region) | Useful when |
|---|---|
| Apex Coffee Imports (US) | Want to buy a Frinsa lot already vetted; pay USD premium |
| Cafe Imports (US) | Broader Indonesian menu |
| Falcon Specialty (UK) | Some Frinsa, Klasik Beans |
| Nordic Approach (NO) | Premium-end micro-lots, expensive |
| Trabocca (NL) | Strong Java + Aceh portfolio |
| Mercanta (UK) | Wide Indonesian menu, container-friendly |
| Local Singapore re-importers (Black Gold, Nylon partners) | Faster freight to MY, USD price + freight markup |
3. Producer Evaluation Framework
When evaluating a new producer or lot, score across these 7 dimensions. Buy only if score ≥ 23/35.
| Dimension | What we look for | Score 1–5 |
|---|---|---|
| 1. Process transparency | Producer can describe their exact fermentation protocol — time, vessel, additions, temperature | |
| 2. Cup quality (sample) | Internal cup score ≥ 3/4 on first sample roast (see Roasting Profiles) | |
| 3. Reproducibility | Producer has produced this lot/process before — not a one-off experiment | |
| 4. Pricing fairness | Within ±15 % of comparable benchmark; producer can explain pricing | |
| 5. Communication | Responds to messages within 5 days; willing to share farm/process photos | |
| 6. Logistics feasibility | Can ship to KL via reasonable channel (direct, exporter, or importer) | |
| 7. Alignment with brand | Story holds up to scrutiny; we’d be proud to show our customers their farm |
If any single score = 1, do not buy. Even a perfect 4 in the others can’t compensate for one fatal weakness.
4. Sample Protocol (the gate before purchase)
Step 1: Identify candidate lot (via direct contact, exporter, or importer).
Step 2: Request 250–500 g sample. Pay for it (don't expect free; it signals seriousness).
Step 3: Sample roast 1 (Profile A or D as starting point). Cup 24–72 h later.
Step 4: If cup score ≥ 2/4, do sample roast 2 with adjustments. Cup.
Step 5: If sample roast 2 scores ≥ 3/4, do sample roast 3 to confirm reproducibility.
Step 6: Score on Producer Evaluation Framework (§3 above).
Step 7: If ≥ 23/35, place a small first order (5–10 kg). Confirm contracted price + ETA.
Step 8: Re-cup the production lot when it arrives. Compare to sample roast 3. Variance > 10 % → contact supplier.
Total elapsed time: ~4–6 weeks from sample request to first production lot. Plan accordingly.
5. Sourcing Calendar (aligned to July 2026 launch)
May 2026 — Outreach & first samples
- Send introductory emails to 6 Tier-S/A producers
- Order 3–4 samples (target: 1 Frinsa, 1 Klasik Beans, 1 Aceh, 1 Bali CM)
- Set up first cupping session at CoRoasting (using their sample roaster)
June 2026 — Production sample roasts + first commit
- Sample-roast all received samples (Roasts 1, 2, 3 each)
- Score on framework
- Place first production order: ~10–15 kg total across 2–3 lots (signature + house + back-up)
- Soft-launch pop-up uses these lots
July 2026 — Launch + lot 2 sample cycle
- Public launch with 2–3 lots in rotation
- Continue sampling: 2 new producers per month
- Place lot 2 production order (15–20 kg)
Aug–Sep 2026 — Establish recurring buy
- Lock recurring Frinsa relationship if validated (signature anchor)
- Add second Aceh micro-lot
- Trial first Sulawesi or Flores lot
Q4 2026 — Forward-buy first harvest
- Aceh harvest (Oct–Jan) → forward-buy 30–50 kg of first-pick lots
- Begin sampling Q1 2027 lots
6. Buying Volumes (year 1 plan)
| Month | Target green inventory (kg) | Notes |
|---|---|---|
| Jun 2026 | 12–18 | Soft-launch + first month buffer |
| Jul | 18–25 | Public launch |
| Aug | 25–35 | Building momentum |
| Sep | 30–40 | Pre-Aceh-harvest buffer |
| Oct | 40–50 | Aceh harvest forward-buy |
| Nov | 35–45 | Roast through stock |
| Dec | 30–40 | Festive demand, year-end |
| Jan 2027 | 40–50 | Aceh fresh-crop hits |
| Feb | 35–45 | Steady state |
Tip
Target steady-state = 1.5–2× monthly roast volume on hand at all times. Less = stockout risk; more = freshness loss.
Cross-ref: Green Inventory Math.
7. Cost Benchmarks (FOB, 2026)
Indicative only. Actual prices fluctuate with FX, harvest, grade.
| Origin / Process | FOB (USD/kg) | RM equivalent (~RM 4.7/USD) |
|---|---|---|
| Aceh Gayo washed (Grade 1) | 8–12 | 38–56 |
| Frinsa anaerobic natural | 14–22 | 66–103 |
| Frinsa fruit co-ferment (signature) | 18–28 | 85–132 |
| Klasik Beans yeast-inoculated | 16–24 | 75–113 |
| Bali Kintamani CM | 14–20 | 66–94 |
| Toraja washed | 10–14 | 47–66 |
| Landed cost in KL (add freight, duty, handling) | +30–50 % | +RM 12–35/kg |
So a Frinsa fruit co-ferment lands at roughly RM 110–170/kg green. After 13 % roast loss → RM 127–195/kg roasted. Per cup (15 g) → RM 1.90–2.93. → see Unit Economics.
8. Sourcing Channels — Comparison
| Channel | Lead time | Min order | Margin / cost edge | Relationship depth | Best for |
|---|---|---|---|---|---|
| Direct from producer | 8–12 weeks | 5–30 kg | Best price | Highest | Recurring signature lots |
| Indonesian exporter (PT-X) | 4–6 weeks | 10–60 kg | Mid | Medium | Quick-turn lots; multi-origin baskets |
| International importer (US/EU) | 6–10 weeks | 10–35 kg (or pre-shipped) | Lower margin (FX + freight) | Low | Specific lots not available domestically |
| SG re-import partner | 1–2 weeks | 5–20 kg | Lower margin | Low | Emergency buys; small testing |
9. Sample Vetting Email Template (Bahasa + English)
Subject: Pertanyaan sample co-ferment / Co-ferment sample inquiry — KNNO Coffee (Malaysia)
Halo [Producer name],
Saya [Founder name], dari KNNO Coffee — sebuah brand kopi pop-up di Malaysia yang spesialisasi pada kopi co-fermented dari Indonesia.
Hi [Producer name],
I'm [Founder name] from KNNO Coffee, a pop-up coffee brand in Malaysia specialising in co-fermented Indonesian coffees. Our menu launches July 2026 in Kuala Lumpur, and we're building a small portfolio of producer-named Indonesian lots.
We'd love to sample your [process / lot name] from the [year] harvest. Specifically interested in:
- 250–500 g green sample (we will pay for sample + shipping)
- Process notes (fermentation duration, vessel, any inoculation, drying)
- Approximate FOB price for a 10–30 kg first order
- Earliest available shipping window
We light-roast for clarity and serve to a tea-drinking audience that's just discovering co-ferments. Your work would be perfectly placed in our story.
Thank you for your time — looking forward to tasting your coffee.
Best,
[Founder name]
KNNO Coffee
[email] · [WhatsApp]
[Instagram handle]
10. Sample Tracking Sheet (use for every sample)
SAMPLE LOG
Sample ID: _________________
Received: __________ Producer: ____________________
Origin: ____________________ Process: ____________________
Harvest year: ______ Variety: ____________________
Sample size: ______ g Cost paid: RM _____ Shipping: RM _____
ROAST 1
Profile used: _____ Cup score: ___/4 Notes: ___________________________
ROAST 2 (adjusted)
Profile adjustment: _____ Cup score: ___/4 Notes: ___________________
ROAST 3 (reproducibility)
Cup score: ___/4 Variance vs. R2: ___ Notes: __________________________
PRODUCER EVALUATION FRAMEWORK SCORE
1. Process transparency: ___/5
2. Cup quality: ___/5
3. Reproducibility: ___/5
4. Pricing fairness: ___/5
5. Communication: ___/5
6. Logistics: ___/5
7. Brand alignment: ___/5
TOTAL: ___/35 Decision: BUY / DECLINE / RE-SAMPLE
If BUY: order qty ______ kg Price/kg RM ______ ETA ______
11. Risk-Specific Sourcing Notes
| Risk | Mitigation in sourcing |
|---|---|
| Single producer over-dependence | No single producer > 40 % of annual green volume in year 1 |
| Harvest failure | 3-region portfolio (Aceh + Java + Bali/Sulawesi) |
| FX shock (MYR/IDR/USD) | Forward-buy 3-month inventory when MYR is favourable |
| Freight delays | Maintain 6+ weeks buffer stock; have SG re-import as fast-channel backup |
| Quality drift between sample and production lot | Always re-cup production lot vs. sample roast 3; variance > 10 % → contact supplier |
| Co-ferment fad burns out | Maintain at least 1 high-quality washed lot at all times (fall-back menu anchor) |
| Producer raises prices abruptly | Multi-producer portfolio; pre-negotiated price ceiling annually |
→ Cross-ref 07-Risks & Failure Modes.
12. Documentation Standards
For every lot we buy and serve, we keep on file:
- Producer name, contact, location (precise: village / cooperative)
- Variety / cultivar
- Altitude, harvest year/window
- Process: detailed (fermentation duration, vessel, additions, drying)
- FOB price + landed cost
- Sample roast log + cupping notes
- Production roast log + cupping notes
- Date received in MY, date roasted, date served / drop window
This data lives in Lot Master File.
My Notes & Thoughts
- The hardest part is patience: the sample → roast → cup → roast → cup → buy cycle takes 4–6 weeks. Start NOW for July launch.
- Frinsa Estate is the obvious “anchor” producer — start there, but don’t make it your only producer in year 1. Two producers minimum by Q4 2026.
- Bahasa-language email gets ~3× the reply rate from Indonesian producers. Use it.
- When you visit Indonesia (year 2, ideally), prioritise producers you’ve already bought from twice. The relationship pays back exponentially.