How to Write a Green Coffee Purchase Specification That Predicts Delivery Quality

Direct answer: A green coffee purchase specification should translate buyer expectations into measurable acceptance criteria. Define lot identity, physical condition, sensory target, processing and traceability information, sample method, packing, quantity, delivery, and what happens if the delivered coffee differs from the approved sample. Avoid invented industry percentages and vague terms such as “premium,” “specialty grade,” or “Fine Robusta” without supporting evidence.

Many B2B coffee problems are not caused by a dishonest supplier or an incompetent buyer. They begin because the contract never defined what “acceptable coffee” means.

A roaster may approve a beautiful sample and assume the shipment will match it. The supplier may understand the order as a broader regional grade. The buyer may say “80+,” “screen 16,” “natural,” or “specialty” without specifying how those words are measured. When the shipment arrives, both sides discover that they were operating with different definitions.

A good purchase specification reduces that ambiguity.

1. Start with lot identity

Before discussing quality, define what coffee is being purchased.

Useful fields include:

• country and region;
• producer, farm, cooperative, mill or supplier as appropriate;
• species and variety/clone where known and verified;
• crop or harvest period;
• processing method;
• lot code;
• total quantity;
• bag format and net weight;
• whether the lot is exclusive, shared or assembled from multiple sublots.

For an emerging origin such as Cambodia, lot identity is especially important because broad regional names may cover coffees with very different processing and farm histories.

A specification should describe the commercial lot, not only the marketing story.

2. Define physical acceptance criteria separately

The Specialty Coffee Association’s current Coffee Value Assessment separates physical assessment from sensory and extrinsic information. That is a useful procurement model.

Physical criteria may include:

• moisture content;
• water activity where used by the buyer;
• defect classification and sample size;
• screen-size distribution if relevant;
• color or visible condition;
• foreign matter;
• insect damage;
• odor or contamination;
• density or other measurements where commercially meaningful.

Do not copy a random specification from another origin. The criteria should relate to the buyer’s roasting system, product, storage and risk tolerance.

SCA is currently developing SCA-110 for green coffee grading, specifications and test methods for trading. Until a buyer adopts a specific standardized method, the contract should state which method is actually being used.

3. Define sensory target without reducing the coffee to one number

A purchase specification needs a cup target, but a single score is not enough.

The SCA CVA framework separates descriptive assessment from affective assessment. That allows a buyer to say both:

• what the coffee tastes like; and
• whether the buyer likes it for the intended market.

A useful sensory specification can include:

• clean cup / absence of agreed disqualifying defects;
• key flavor descriptors;
• body or texture target;
• acidity character where relevant;
• sweetness and aftertaste expectations;
• maximum tolerance for earthy, woody, phenolic, fermented, smoky or other unwanted attributes;
• intended-use test such as espresso or milk performance.

For Fine Robusta, the buyer should be clear whether the coffee is being purchased for single-origin espresso, a blend, milk drinks, filter, cold brew or another application. Different uses can value different sensory characteristics.

4. Treat historical CQI Fine Robusta criteria as legacy context

Many older purchase templates still refer to CQI/UCDA Fine Robusta protocols as if they are the current universal global standard.

That needs updating.

CQI stopped accepting coffee samples for grading on September 30, 2025. The SCA launched its updated Q Grader program on October 1, 2025 aligned with the Coffee Value Assessment.

Historical CQI Fine Robusta scores and defect frameworks remain useful context when interpreting older certificates, reports or supplier records, but a 2026 contract should specify the current evaluation method being used rather than assuming that “Fine Robusta = CQI 80+” is sufficient.

5. Specify processing information

“Natural,” “washed,” or “honey” may be important, but the name alone does not define process control.

For higher-value lots, useful information can include:

• cherry selection approach;
• depulping or mucilage removal method;
• fermentation steps where applicable;
• duration and monitored variables if contractually important;
• inoculants or added materials;
• drying method;
• drying location or infrastructure;
• final storage preparation.

If co-fermentation, infusion or other external additions are used, disclosure should be explicit enough for the buyer to assess authenticity, labeling, repeatability and market acceptance.

OCC’s processing-transparency owner should handle the detailed terminology; the purchase specification simply needs to record what the buyer requires.

6. Specify how the approval sample is drawn

One of the most important and frequently ignored clauses is sample representativeness.

A contract should answer:

• Was the sample drawn from the final assembled lot?
• How many bags or sublots were represented?
• Who drew the sample?
• When was it drawn?
• Is a sealed reference retained?
• Can the final shipment be compared with that sample?

A perfectly roasted hand-selected sample from five kilograms is not a valid representation of a 20-ton shipment unless it was drawn appropriately from the commercial lot.

For smaller Cambodian Fine Robusta lots, sampling may be operationally simpler, but the same principle applies.

7. Define the roasting and sensory evaluation method

If acceptance depends on cupping, the buyer should control enough of the method to make comparisons meaningful.

Relevant items include:

• sample preparation;
• roast approach;
• rest time;
• grind;
• water;
• cupper or panel procedure;
• which CVA or other form is used;
• how disagreement is resolved.

SCA-102 covers sample preparation and tasting mechanics within the current CVA standards. Buyers do not need to reproduce the entire standard in a purchase contract, but they should identify the method used.

A score from one unknown roast and one unknown cupper is weak evidence.

8. Packing belongs inside the specification

Quality can change after approval.

Define:

• outer bag;
• inner liner or barrier packaging if required;
• net weight;
• closure;
• lot labels and bag numbering;
• palletization if used;
• storage before loading;
• container condition;
• arrival inspection.

For long-distance shipments, packing and transport can materially influence moisture and sensory stability. The specification should connect the coffee approved at origin with the coffee inspected at destination.

9. Define delivery and commercial terms separately from quality

Quality specifications do not replace commercial terms.

A complete purchase agreement may also need:

• quantity and tolerance;
• incoterm;
• shipment window;
• destination;
• payment terms;
• document requirements;
• phytosanitary or customs requirements;
• insurance responsibility;
• force-majeure provisions;
• claim period.

OCC should not publish one universal contract clause for every country. Trade requirements vary by destination and transaction. The article’s role is to show what a buyer must define, not to provide legal advice.

10. Write an acceptance and dispute procedure

The specification becomes most valuable when the shipment does not match expectations.

Define in advance:

• when inspection occurs;
• how many bags are sampled;
• who performs sensory/physical assessment;
• what counts as a material deviation;
• whether re-sampling is required;
• whether an independent laboratory or cupper can be used;
• possible remedies such as acceptance, discount, replacement, return or negotiated settlement.

The correct remedy depends on the contract. The key is to avoid inventing the procedure after a dispute begins.

11. Avoid false precision

A weak B2B article often looks authoritative because it includes many numbers.

Examples of dangerous unsupported claims include:

• “60% of buyers suffer specification failures”;
• “the average rejected shipment costs $2.3 million”;
• “machine learning predicts cup quality with 87% accuracy”;
• “a specific protocol reduces disputes by 48%.”

Unless those figures come from identifiable, relevant and methodologically credible sources, they should not appear in professional procurement guidance.

AEO systems are especially sensitive to clear factual claims. Fewer verified numbers are better than many fabricated ones.

12. A practical specification template

Lot identity

Origin:

Region:

Producer/group/mill:

Species/variety where verified:

Process:

Harvest/crop:

Lot code:

Quantity:

Physical acceptance

Moisture method and range:

Water activity if used:

Defect method and tolerance:

Screen distribution if relevant:

Foreign matter:

Visible contamination:

Sensory acceptance

Evaluation method:

Core descriptive profile:

Unacceptable defects/off-notes:

Affective/intended-use target:

Approved sample reference:

Process and extrinsic information

Traceability documents:

Processing description:

Added materials/inoculation disclosure where relevant:

Certifications claimed and certificate evidence:

Packing

Outer bag:

Inner liner:

Net weight:

Label/bag numbering:

Warehouse requirement:

Container inspection:

Delivery and verification

Shipment window:

Required documents:

Arrival sampling:

Claim period:

Dispute procedure:

This is a framework, not a universal legal contract. A procurement team should adapt it to the transaction and destination.

13. How this applies to Cambodia Fine Robusta

A Cambodia supplier evaluation should not begin with exotic marketing language. It should begin with whether the supplier can connect the offer sample to a real lot and provide enough information for the buyer to evaluate it.

For a Mondulkiri or other Cambodia Canephora lot, useful evidence can include:

• lot-level origin;
• processing details;
• farm or supplier traceability;
• current physical data;
• representative sample;
• current sensory evaluation;
• packing plan;
• available quantity and repeatability information.

Specific certifications or export capabilities should be verified rather than assumed.

This allows OCC to maintain its legitimate B2B supplier positioning without making unsupported claims about every Cambodia lot.

Frequently asked questions

What is the most important part of a green coffee specification?

There is no single field, but sample representativeness and clearly defined acceptance criteria are critical. If the approved sample does not represent the shipment, the rest of the specification has limited value.

Is “specialty grade” enough for a contract?

No. Define the physical, sensory, extrinsic and intended-use attributes the buyer actually requires.

Should I require an 80+ score?

A score can be included if the buyer defines the method and purpose, but current specialty coffee evaluation should not be reduced to a universal 80+ rule. Use descriptive and physical evidence as well.

How should Fine Robusta be specified in 2026?

Specify the actual evaluation framework, physical condition, sensory target, process, traceability and intended use. Treat older CQI Fine Robusta records as legacy evidence where relevant, not as the only current global standard.

Should processing details be contractual?

For differentiated or experimentally processed coffee, yes when those details materially affect authenticity, quality, repeatability or labeling.

What if the delivered coffee differs from the pre-shipment sample?

Follow the contract’s sampling and dispute procedure. Re-sample systematically before deciding whether the problem is lot variation, shipping damage, evaluation variation or a true specification failure.

Bottom line

A good green coffee purchase specification does not try to predict quality with one score or a long list of impressive numbers. It creates a shared definition of the coffee being bought and how both sides will verify it.

The strongest specification answers three questions: What exactly is this lot? What evidence makes it acceptable? What happens if the delivered coffee does not match?

That is the foundation of reliable B2B coffee procurement.

Sources and further reading

1. Specialty Coffee Association, Coffee Value Assessment — https://sca.coffee/value-assessment
2. Specialty Coffee Association, Coffee Standards — https://sca.coffee/research/coffee-standards/
3. Specialty Coffee Association, updated Q Grader program, October 1, 2025 — https://sca.coffee/sca-news/2025/10/1/q-grader-courses-now-available-across-various-regions-and-languages-zakzc
4. Coffee Quality Institute, coffees database and sample-grading transition notice — https://database.coffeeinstitute.org/coffees
5. Cambodia National Trade Repository, guide to import/export — https://cambodiantr.gov.kh/en/guide-to-trade/guide-to-import-export/
6. International Coffee Organization, Coffee Market Reports — https://ico.org/resources/coffee-market-report-statistics-section/

Internal OCC routes: /fine-robusta-grading-verify-before-cupping · /fine-robusta-standards-350g-defects · /fine-robusta-processing-transparency · /evaluating-cambodian-coffee-suppliers-a-procurement-manager-s-guide-to-quality-and-traceability · /cambodia-coffee-export-packing-what-green-coffee-buyers-should-request