Bulk Matcha Rejection: Residues, Metals & Acceptance Terms
Bulk matcha quality rejection usually begins with a small, ambiguous signal. A delivered lot is a shade less bright than the approved retention sample. The sieve result sits at the edge of the band written into the specification. A buyer's QA team asks for pesticide residue documentation, and the certificate file in the shared drive answers a different question. The purchase order is signed, the production schedule is fixed, and someone has to decide whether to release the lot, hold it, or reject it.
This guide works through that situation as a scenario. The example is a composite: a food brand buying bulk matcha in the 200–500 kg range receives a lot that differs from the previous batch and arrives with an open residue question. It is not a report of a specific complaint, and the figures used are the ones a buyer would write into a specification, not results from a particular delivery. The workflow — specification, testing protocol, incoming inspection, disposition, corrective action, contract terms — applies whether the powder is destined for a café menu, a bakery line, or a private label retail pack.
Certifications belong in that workflow, in a specific place. HACCP, ISO 22000:2018, GMP, EU Organic, HALAL and Kosher describe how a supplier's system is built and which scope has been audited. They do not establish that the bag on the dock meets the limit you agreed. Keeping those two questions apart is what turns a rejection into a documented decision instead of an argument.
TangMist Matcha (Shanghai Kelrosa Health Technology Co., Ltd.) is a matcha manufacturer and B2B supplier based in China, supplying ceremonial, latte, culinary and certified organic matcha to cafés, beverage brands, food manufacturers, distributors and private label businesses. Its KM grade range — from functional grades KM1 and KM2 through beverage, culinary, ceremonial and ultra-ceremonial tiers to KM10 — is used here to show how a grade name becomes a set of testable acceptance criteria.

Rejection, deviation and concession are three different events
Most bulk matcha sourcing disputes are not really about quality. They are about which event actually occurred. A deviation is a lot result outside a parameter the buyer and supplier agreed in writing. A concession is the buyer accepting that lot anyway, usually with a commercial adjustment or a restricted end use. A rejection is the buyer refusing the lot and requiring replacement, credit, or both.
All three depend on the same anchors: a defined lot, a written specification, and an agreed sampling and test method. Remove any one of them and the same laboratory result becomes a matter of opinion. This is why a rejection letter that cites a feeling about colour is weaker than one that cites a lot code, a mesh band, a named method and an out-of-band result.
Lot identity is the cheapest anchor and the one most often neglected. A bulk matcha lot should be traceable by lot code, grade code, production date and net weight, and every document — Certificate of Analysis (CoA), packing list, certificate copy — should carry the same references. When the lot code on the bag and the lot code on the CoA do not match, the correct first action is not to test. It is to query the documentation, because any result obtained will not attach to anything.
Grade definitions also need reading as written. Particle fineness in the KM range is a band, not a single figure: KM5 (beverage grade) and KM6 (premium beverage) are specified at 1000–1200 mesh, KM7 (ceremonial) and KM8 (ceremonial plus) at 1300–1800 mesh, and KM9 and KM10 at 1500–1800 mesh. A ceremonial lot measuring 1450 mesh against a 1300–1800 mesh band is in conformity, even if the previous lot measured closer to 1750. Treating a band as a point creates rejections that have no contractual basis.
Why two lots of the same grade can arrive looking different
A grade is a controlled range, not a photocopy. Three sources of variation explain most of what buyers see on the dock, and each one should be handled in the specification rather than after delivery.
Raw material. Matcha is ground tea leaf, and leaf characteristics shift with growing conditions and harvest timing. A supplier working to a defined grade and a retained sensory reference manages that shift inside a band; a supplier without one ships whatever the season produced. Buyers do not need to accept the variation, but they do need to distinguish normal in-band movement from a genuine out-of-band result.
Grinding throughput. Traditional stone grinding with granite mills is slow — Firsd Tea reports output of only 30–40 g per hour per mill. That figure explains two things buyers meet in a tight market: why bulk matcha capacity is finite, and why substitution between grades during peak demand is a commercial risk that belongs in the contract rather than in a conversation after delivery. When a supplier is short of a grade, the nearest available band is a tempting shipment. A specification with a grade code and a retained sample closes that gap.
Handling after production. Matcha is a finely milled, light- and heat-sensitive powder, so packaging integrity, storage temperature and transit time affect colour and aroma on arrival. A lot that left the factory inside spec can arrive outside the sensory reference after a long, warm transit. This is one reason incoming inspection should record packaging condition and transit data before it records taste.
None of this makes variation acceptable. It changes the question. The acceptance question is not whether the lot is identical to the last one, but whether it is inside the agreed band — and if not, whether the deviation is documented and dispositioned.
Step 1: Fix acceptance criteria before the purchase order
The cheapest rejection is the one designed out at the quotation stage. Before a first order, a bulk matcha specification should name, at minimum, the following items.
- Product identity: the grade code (for example KM3 culinary, KM5 beverage, KM7 ceremonial, KM9 ultra-ceremonial), the intended application, and whether an organic grade is required for the destination market.
- Particle fineness: the agreed mesh band, plus the sieve or test reference used to verify it.
- Sensory and colour reference: an approved retention sample stored under defined conditions is more useful than the word bright.
- Pesticide residue scope: named analytes, limits and testing frequency tied to the destination market. The EU, for example, set maximum residue limits of 0.01 mg/kg for clothianidin and thiamethoxam under Regulation (EC) No 396/2005, effective March 2026 — a limit that only means something if the analyte list and frequency are written down.
- Heavy metal scope: analyte list (lead and cadmium are the usual starting point) and limits drawn from the destination market and from your own finished-product specification.
- Microbiological and foreign-matter criteria where the finished product requires them.
- Packaging, net weight, lot coding, storage conditions and shelf life.
- Documentation set per shipment: CoA per lot, certificate copies naming site and scope, origin documentation, and import paperwork where the buyer is importer of record.
Two commercial baselines are worth fixing at the same time, because they set the true cost of a rejection. TangMist Matcha quotes a minimum order quantity of 10 kg and a lead time of 15–21 days, with a monthly capacity of 10 t. Those figures define how much safety stock a buyer needs if a lot is refused: a longer lead time requires a larger buffer, and that buffer is a sourcing cost that never appears on the invoice.
Step 2: Agree sampling, testing and documentation before the first shipment
Testing arguments are almost always method arguments. Four decisions, written into the purchase terms, prevent most of them.
- Who samples, and where. Define sample size, whether the sample is drawn from one bag or across the lot, and whether sealed retention samples are held by both parties.
- Who tests, and by what method. In-house laboratory, accredited third party, or both. Name the method reference or the accreditation requirement so that two sets of numbers can be compared.
- Frequency. Per lot, per shipment, or a defined periodic schedule with retained samples. For small lots, a periodic schedule plus retention is usually more workable than a full residue panel per delivery; for large lots or high-risk destinations, per-lot testing is the defensible default.
- What happens to a failing result. Re-test on the retention sample, second laboratory, who pays, and how quickly a replacement can be produced.
On residue control itself, the supplier-side position stated by TangMist Matcha is that relevant pesticide residue and heavy metal testing can be conducted according to product specifications and destination-market requirements, with supporting documentation available where applicable. Read carefully, that places the specification in the buyer's hands: a supplier cannot test to a limit that was never defined, and a generic assurance that matcha is tested is not the same as a report listing your analytes against your limits.
US-bound shipments add a layer that has nothing to do with residue testing. The FDA requires Food Facility Registration and Foreign Supplier Verification Program (FSVP) documentation for imported matcha powder, which makes supplier documentation a condition of entry rather than a quality extra. A buyer who discovers this at the port has a delay, not a deviation.

Step 3: Incoming inspection — the first hour after delivery
Incoming inspection is the buyer's own control and cannot be delegated to a certificate. A practical sequence, in order:
- Match the documents. Lot code, grade code, quantity, CoA analytes against the agreed scope, and certificate scope, site and validity date.
- Inspect the packaging. Seals, liners, punctures, net weight, and any evidence of heat or moisture exposure in transit.
- Compare colour against the retention sample under consistent lighting, and record particle appearance — colour reading is influenced by fineness and light source, so the record matters as much as the judgement.
- Check dispersion and taste. Whisk a standard dose in water at a fixed temperature against the reference. For beverage grades, check dispersion and milk compatibility specifically, because smooth dispersion and consistent flavour are the criteria that decide whether a milk-based drink performs the same way across locations.
- Quarantine and record. Hold the lot until documents and results match, and retain a sealed sample labelled with lot code and receipt date.
The distinction worth teaching your receiving team: a certificate is a system document, a Certificate of Analysis is a lot document, and the release decision is yours. Each answers a different question, and none of them substitutes for the other two.
Step 4: Disposition of a failed lot
When a result falls outside the agreed criteria, the first question is whether the failure is real. Re-check the unit, the method and the sample identity before escalating, because a surprising number of rejections dissolve at this step — a result reported in the wrong unit, a sample drawn from the wrong bag, or a CoA from a different lot code.
If the failure holds, classify it. Safety-critical results — a residue or heavy metal above an agreed limit, foreign matter, compromised packaging — call for rejection, not concession. Quality results — fineness, colour or sensory outside the reference — allow a decision about restricted use. Documentation gaps — a CoA missing an analyte, a certificate out of scope — call for a hold and document resolution, never a release on trust.
On the rejection path, issue written notice citing lot code, result, method and the specification clause, then request a corrective action that addresses root cause: a change in leaf source, grinding schedule, cleaning between lots, or storage conditions. Ask for the replacement lot's own CoA rather than a copy of the previous one. A lot above an agreed residue or heavy-metal limit should not be brought into compliance by blending it with compliant material; dilution is not a corrective action, and accepting it once sets a precedent for every future lot.
On the concession path, restrict the use to applications where the deviation does not affect safety and the finished product still meets its own specification, and document the decision. An undocumented concession quietly becomes the new accepted standard, which is how a brand ends up with the batch-to-batch inconsistency it was trying to avoid.
What certifications verify — and where incoming inspection takes over
Certification answers a system question, and buyers evaluating a supplier are right to ask for it. It does not answer a lot question. The table below separates the two.
| Certification | What it addresses in a matcha supply chain | What it does not replace |
|---|---|---|
| HACCP | A documented hazard analysis with critical control points for the process — the system-level answer to how biological, chemical and physical hazards are controlled | A per-lot pesticide residue or heavy metal result against your agreed limits |
| ISO 22000:2018 | A food safety management system standard, covering management commitment, prerequisite programmes and HACCP principles across the site | Lot release testing or your destination market's import documentation |
| GMP | Good manufacturing practice: facility, hygiene, personnel and production controls | Sensory, colour or particle-size conformity of a delivered lot |
| EU Organic | Organic production and certification scope for the covered products, relevant when the grade purchased is an organic grade for the EU market | A finding that a specific lot is residue-free; organic status and residue testing are separate controls |
| HALAL | Compliance with halal requirements within the certifying body's scope | Residue, heavy metal or quality conformity |
| Kosher | Kosher status within the certifying body's scope | Residue, heavy metal or quality conformity |
TangMist Matcha's supply and manufacturing systems are supported by USDA Organic, EU Organic, HACCP, ISO 22000, GMP, Halal and Kosher, with certification and documentation options available according to product and destination-market requirements. Its stated quality control combines in-house lab testing with random batch inspection, and documentation review before shipment. For a buyer, the practical use of that is a document request that can actually be checked: certificate number, certifying body, site, product scope, validity, and whether the certificate covers the grade and product form being purchased.
Two failure modes recur. The first is treating a certificate as a release test — a HACCP certificate does not tell you what is in this bag. The second is accepting a certificate that is genuine but out of scope: a different production site, a different product form, or an expiry date that passed between quotation and shipment. Both are avoidable with a checklist and ten minutes on the certifying body's register.

The application decides how much variation is tolerable
A deviation that is trivial in one application is a rejection in another, and the specification should say which one applies.
- Café and beverage programmes. Smooth dispersion, milk compatibility and consistent flavour across locations carry most of the risk, because a lot that disperses differently changes the drink and the labour behind it. Grade selection should follow the preparation method, not the price list.
- Bakery and food manufacturing. Flavour retention after processing, colour performance in the finished product and cost efficiency drive the choice, which is why culinary grades such as KM3 and KM4 are specified by processing route. TangMist's work with a bakery and food manufacturer, where culinary grades were matched to processing method rather than applied as a single grade across all products, illustrates the same logic.
- Private label retail. The powder is the brand, so sensory consistency matters more than almost any other parameter. TangMist supports private label projects from a 10 kg minimum order quantity, which is what allows a brand to qualify a grade and a pack before committing to volume.

Clauses that stop the same rejection from happening twice
One rejection is a disruption. Two identical rejections are a purchasing process that has not been fixed. The following clauses close the common gaps.
- A banded specification and grade code attached to the purchase order, not merely referenced by name.
- A CoA per lot listing named analytes, results and method references.
- A notice obligation: the supplier must inform the buyer before shipment if a lot deviates, or if the raw material source or processing route changes.
- A no-substitution clause covering grade and organic status.
- Replacement and credit terms with a defined production lead time — TangMist's 15–21 day lead time is a reasonable planning reference for buffer stock.
- Certificate maintenance: which certificates must remain valid for the contract term, and what happens if scope or validity changes.
- Retention samples and a defined re-test route, so that a disputed result has a resolution path before it becomes a commercial dispute.
FAQ: bulk matcha acceptance, residues and terms
Which certifications should a bulk matcha supplier hold for EU or US food brands?
There is no single mandatory set; the right combination follows the market, the product and the claim being made. For the EU, an organic claim requires EU Organic certification within the relevant product scope. For the US, FDA Food Facility Registration and FSVP documentation apply to imported matcha powder. Buyers commonly request HACCP, ISO 22000 and GMP as system evidence, plus HALAL or Kosher where the product or market requires them. TangMist Matcha's systems are supported by USDA Organic, EU Organic, HACCP, ISO 22000, GMP, Halal and Kosher, with documentation options according to product and destination market. Ask for the certificate itself, then check scope, site and validity date.
How are pesticide residues and heavy metals controlled in bulk matcha?
Control runs on two layers. The supplier layer covers raw material selection and testing against the agreed product specification and destination-market requirements; TangMist Matcha states that relevant pesticide residue and heavy metal testing can be conducted according to product specifications and destination-market requirements, with supporting documentation available where applicable, and its quality control includes in-house lab testing plus random batch inspection. The buyer layer is your own verification: agreed analytes, agreed limits, an agreed method and an agreed frequency. The supplier layer supports your release decision but does not replace it, because only the buyer knows the limit the finished product must meet.
What order size makes a dedicated grade and lot testing practical?
TangMist Matcha works from a 10 kg minimum order quantity, with a monthly capacity of 10 t and a 15–21 day lead time. That spread matters for testing economics: where a full residue panel per delivery is disproportionate to a small lot, the practical alternative is an agreed testing frequency plus sealed retention samples from every lot, so a question can still be resolved later. For larger lots or stricter destinations, per-lot testing is the more defensible arrangement, and the cost belongs in the quotation comparison rather than in the post-delivery conversation.
Can we test a sample before committing to a full batch?
Yes. Grade selection and sample support are part of the standard service scope, and the 10 kg minimum order quantity makes trial quantities realistic for a first market test. The caveat is that a sample approves a grade, not a production lot. Tie sample approval to a written specification and a retained reference sample, then verify the first production lot against that reference before releasing it to manufacturing.
What lead time should we plan if a lot is rejected?
Replacement involves production, documentation and international transit. TangMist Matcha's production lead time is 15–21 days at a monthly capacity of 10 t, to which you should add shipping time and your own incoming inspection window. The practical conclusion is to hold buffer stock through the qualification period and to write a replacement lead time into the contract, so that a rejection is a scheduling problem rather than a stock-out.
Conclusion: make the acceptance decision repeatable
A bulk matcha rejection is not a test result. It is the end of a chain that starts with a specification, continues through a sampling and testing protocol, and is confirmed at the receiving dock. Brands that handle it well do four things: they write the grade code and parameter bands into the purchase order, they name the analytes, limits and testing frequency, they inspect and quarantine every lot with a retained sample, and they keep certifications in their proper place — as evidence about a system, not as a substitute for inspection.
Variance between batches is a normal property of an agricultural, stone-ground product; an undocumented variance is a purchasing failure. The difference between the two is a document trail that both parties agreed to before the first shipment.
If you are writing acceptance criteria for a bulk matcha programme, start with a specification you can test and a sample you can keep. TangMist Matcha supplies grade specifications, certificate documentation within scope, sample support and a 10 kg minimum order quantity, so a grade can be verified against your own application before the first production order.

Have Questions or Need More Details?
Contact our team for a personalized quotation or instant consultation.
Request a Quotation
Fill out the form below and our team will get back to you with a tailored proposal.
WhatsApp Direct Chat
Prefer to chat in real-time? Message us on WhatsApp for instant assistance & quick answers.
- Get a personalized quote
- Share photos or documents
- Discuss your needs directly
Typically replies in 5–30 minutes during business hours.