How to Verify an Indian Spice Exporter Before You Pay

Most buyers who lose money on an Indian spice order do not lose it to an obvious fraud. They lose it to a counterparty who was real, registered, responsive and simply not capable of doing what they said — a trading office that quoted a processor's specification without a processor behind it.

Verification is therefore less about detecting criminals and more about establishing capability. The process below has four layers, in increasing cost and increasing certainty. How far you go should scale with the order value and how much of your own reputation rides on the goods.

Layer one: does the legal entity exist, and is this it?

This layer costs nothing but time and eliminates a surprising proportion of counterparties.

Ask for five things in writing, and then verify each one independently rather than accepting the PDF:

• Import Export Code (IEC) — issued by India's Directorate General of Foreign Trade. Every legal exporter has one.
• FSSAI licence number — with the licence category visible. This matters: FSSAI licences distinguish manufacturers from traders and repackers, so the category itself tells you what the holder is registered to do.
• Spices Board CRES registration — Registration as an Exporter of Spices, mandatory for spice exports from India.
• GSTIN — the goods and services tax identity, which encodes the state and the PAN of the entity.
• Company registration — CIN for a private limited company, or the equivalent for an LLP or proprietorship.

What you are checking is not merely that the numbers exist but that they all resolve to the same legal name at the same address, and that the name matches the one on the quotation, the proforma invoice and — critically — the bank account you will eventually be asked to pay.

The most common finding at this layer is not fraud. It is that the trading name on the website is a brand, the IEC belongs to a related entity, and the FSSAI licence is a trader category rather than a manufacturer category. None of that is necessarily disqualifying. All of it is information you should have before you negotiate.

The GSTIN detail worth knowing

An Indian GSTIN embeds the PAN of the registered entity in characters three to twelve. That means the GSTIN and the company's PAN-linked identity should agree. It also means two entities claiming to be the same business but showing GSTINs with different embedded PANs are, in fact, two different businesses. This is a thirty-second check that occasionally saves an entire negotiation.

Layer two: do they process, or do they buy and resell?

This is the layer that most buyers skip, and it is the one that determines whether your specification is enforceable.

A merchant trader can source excellent material. What they cannot do is *hold a specification across lots*, because they do not control the line that produces it. When your second container differs from your first, a trader's honest answer is that the market changed. A processor's honest answer is a batch record.

Four questions, asked in one call, separate them reliably:

1. Do you own the processing facility? Then: is it on the same premises as the registered address, or elsewhere? A different address is fine and common; an evasive answer is not.
2. Walk me through the line, live, right now. Not a scheduled visit, not a video file. A processor will do this from a phone in about six minutes.
3. Which laboratory runs your residue panel, and is it NABL-accredited? A processor names the lab without hesitating because they use it weekly.
4. Show me a Certificate of Analysis from a commercial shipment last month. Not from a sample prepared for you — from a real consignment to a real customer, with the customer's details redacted.

That fourth question is the strongest single discriminator in this entire guide. Sample CoAs are easy to produce. A file of recent commercial CoAs, with consistent parameters across lots, is the fingerprint of an operating quality system.

What a genuine factory walk shows

If you get the live walk, know what you are looking for. In sequence, a spice processing line runs roughly: raw material intake and quarantine, pre-cleaning, destoning and gravity separation, metal detection, sometimes steam sterilisation, grading or sizing, grinding if applicable, final metal detection, weighing and packing, and a finished-goods store.

Ask the person to name each machine as they pass it, and to open a batch record at one of them. Someone standing in a facility they operate does this without breaking stride. Someone standing in a facility they have hired for the afternoon hesitates at the batch record, because the paperwork is the part that cannot be borrowed.

Watch the incidental detail rather than the equipment: whether the floor near the intake is clean, whether staff are wearing what the facility's own signage says they should, whether the finished-goods store has product from other customers in it — that last one is a good sign, not a bad one, because it means the line runs.

Layer three: test what they ship, not what they sample

The gap between sample and shipment is where most quality disputes live, and it is almost always a sampling problem rather than a substitution one.

A sample sent by an exporter is drawn by the exporter, from material the exporter chose. Even with complete good faith, it is the best of what they have. Your protection is to move the point of sampling.

• Appoint the inspector yourself. SGS, Bureau Veritas, Intertek and TUV all operate in Indian spice regions. You appoint, you brief, you pay. An inspection arranged by the seller is a seller's document.
• Specify the sampling plan. Number of bags sampled, whether sampling is random across the lot, how the composite is formed, and how the retained sample is sealed and stored. "A sample was taken" is not a sampling plan.
• Test against your destination's standard, not a generic screen. A multi-residue panel scoped to your regulation, plus aflatoxin where the commodity warrants it, plus the microbiological set.
• Keep a sealed retained sample. Split three ways: exporter, inspector, you. Disputes resolve quickly when there is an agreed physical reference and slowly when there is not.
• Test again on arrival and track the delta. Across three shipments, the difference between origin and arrival results tells you whether the supplier's laboratory is honest, optimistic, or simply sampling badly.

That last practice is the most underrated instrument in the whole trade. It costs almost nothing beyond tests you would run anyway, and it converts a subjective sense of whether a supplier is reliable into a number you can act on.

Layer four: structure payment so that being wrong is survivable

No amount of verification reduces risk to zero. The final layer accepts that and limits the consequences.

Documentary letter of credit. Slower and it carries bank charges, but the conditions are enforced by a bank rather than by relationship. Name the Certificate of Analysis and the independent inspection certificate as required documents; a discrepancy then stops payment automatically rather than requiring you to argue.

Documents against payment. Lighter than an L/C. Title documents move with payment, so the goods do not release without it.

Partial advance against inspection. Twenty to thirty percent on order, the balance against a clean inspection certificate. Reasonable once you have a track record with the counterparty, and considerably better than a full advance.

Open account. Only after several clean shipments, and ideally credit-insured.

Three payment red flags are worth treating as stops rather than negotiating points:

• The bank account name does not match the exporter's legal name. Occasionally legitimate, always to be documented and verified before, not during, payment.
• The account details change late, by email, with urgency. This is the standard shape of both invoice-interception fraud and internal diversion. Verify any change by voice, on a number you already had, not one in the email.
• Pressure for a larger advance than agreed, justified by a market movement. Prices do move. Contracts exist for that reason.

The specific patterns that precede a loss

Fraud in this trade is not creative. The same handful of patterns recur, and each has a tell that appears before the money moves.

The price that is meaningfully below the market. Spice is a commodity with visible reference prices. A quotation ten or fifteen percent under the market for the stated grade is not a bargain; it is a different grade, a different crop year, or a different set of intentions. Ask what specifically makes it cheaper. A legitimate answer exists surprisingly often — an exporter clearing carry-over stock, a favourable position on an earlier purchase — and a legitimate exporter gives it without hesitation.

The company that only exists in one direction. A website, a directory listing and a Gmail address, but no registrations that resolve, no landline, no verifiable address, no trace of having shipped anything. Genuine exporters leave a wider footprint than a marketing page.

The urgency that arrives with the invoice. Vessel closing today, price valid until this evening, container already booked. Manufactured time pressure exists to prevent verification, and it is the most reliable single indicator in this entire guide. A real exporter will hold a price for the two days it takes you to check them, because they would rather have a customer than a deposit.

The sample that is too good. A sample materially better than the grade you asked for is not generosity. It is establishing an expectation that the shipment will not meet, and it is why sampling should move to an inspector you appointed.

The document set that does not reconcile. Company name on the quotation, a slightly different name on the proforma, a third name on the bank details. Each difference explained individually and plausibly. Taken together, a structure designed so that no single entity is accountable.

The reference you cannot contact. References that can only be reached through the exporter, or that respond in the same phrasing as the exporter, are not references.

Proportionality: how much verification is enough

Running all four layers on every order is not sensible. A workable rule:

The one layer never worth skipping is the fourth. Verification tells you what is likely; payment structure decides what happens when the unlikely occurs.

Verifying a trading company you have decided to keep

Not every good relationship is with a processor, and treating "trader" as a slur is a mistake that costs buyers access to genuinely useful counterparties. A good trading house carries inventory, absorbs currency risk, consolidates part-loads, and knows which of thirty processors has capacity this week. Those are real services.

What changes is *what* you verify. With a processor you are verifying a facility. With a trader you are verifying a supply chain and a balance sheet.

Ask for:

• Named upstream processors for the product you are buying, and permission to verify one of them. A trader with nothing to hide will facilitate this; a trader who treats their supplier list as the entire value of the relationship will resist, which is itself informative.
• Evidence of inventory or contracted position. A trader selling you material they do not yet own is taking a market position with your order, and if the market moves against them, your shipment is the one that gets renegotiated.
• Years trading and shipment history in the specific commodity, not in "spices" generally. Cumin and cardamom are different businesses.
• Who carries the quality liability. If the processor's CoA is the only quality document, the trader is a payment intermediary, and you should be verifying the processor.

The workable structure with a trader is usually to verify the processor once, then transact through the trader with the processor named in the contract. You get the trader's logistics and flexibility with the processor's accountability.

What verification cannot tell you

Two honest limits are worth stating, because guides in this space rarely do.

Verification establishes capability at a point in time. It does not predict a bad harvest, a currency move, a change to a destination residue limit, or a key employee leaving. A supplier who was excellent for two years can have a poor season, and the evidence you gathered at onboarding will not warn you.

And no amount of documentation substitutes for testing the actual lot. Registrations, audits and certificates are filters that decide who you talk to. The Certificate of Analysis on the consignment in front of you is the only document that speaks to the goods you are buying.

Auditing the facility properly, when the order justifies it

A live video walk is layer two done cheaply. For a relationship you intend to build volume on, a physical audit is the version that holds up.

You have three routes, and they differ more in what they tell you than in cost.

A third-party audit to a recognised standard. SGS, Bureau Veritas, Intertek, TUV and SAI Global all run food-safety audits in India against BRCGS, FSSC 22000 or a customer-specific protocol. You get a structured report against defined clauses, and a supplier who passes has been measured against the same yardstick as your other suppliers. This is the right choice when the output has to satisfy your own retail customers.

A technical visit by someone who knows the commodity. Less structured, and often more revealing. A person who has processed pepper will notice in ten minutes what an auditor working from a checklist may not: that the destoner is undersized for the throughput being claimed, that the sterilisation cycle logs are all suspiciously identical, that the finished-goods store has no separation between treated and untreated material.

Your own visit. Worth doing once for a strategic supplier, mostly for the relationship rather than the findings. You will not out-audit a professional auditor on a first visit, but you will learn who actually makes decisions, which is not always the person who has been emailing you.

Whichever route, ask for the audit to cover four things specifically, because generic protocols cover them lightly:

1. Traceability. Pick a finished lot at random and ask them to trace it back to raw-material intake, and forward to the customers it shipped to. A working system does this in under an hour. A paper system cannot do it at all.
2. Sterilisation validation. If they claim steam sterilisation, ask for the validation study and the routine cycle records, not the equipment brochure.
3. Metal detection and foreign-body control. Detector sensitivity, challenge-test frequency, and what happens to a rejected pack.
4. Segregation of organic and conventional, if you are buying certified organic. This is the most common failure point in Indian organic chains and the hardest to fix after the fact.

What a passed audit does not cover

An audit assesses the system, on the day, at the site audited. It does not tell you whether the material in your lot came from that site — which matters enormously in a trade where processors routinely job out overflow to third-party lines during peak season. Ask explicitly whether any part of production is subcontracted, and if so, whether the subcontractor is in scope of the audit. The honest answer is often yes and out of scope, and it is far better to know that in advance than to discover it when a lot behaves differently.

Reading a Certificate of Analysis properly

The CoA is the document that decides whether you accept a shipment, and most buyers scan it for pass/fail rather than reading it. Four things are worth checking every time.

Does the sample identity tie to the lot? The CoA should carry a lot or batch number that appears on the packing list and on the bags. A CoA with no traceable lot reference certifies nothing in particular.

Is the laboratory accredited, and for these tests? NABL accreditation is scope-specific. A laboratory accredited for microbiology is not thereby accredited for pesticide residue. The accreditation certificate number and scope should be on the report.

Are the limits the right ones? A report that shows "pass" against Indian domestic limits tells you nothing about whether the goods clear in Hamburg. The report should name the standard it is testing against, and it should be yours.

What is not on the report? This is the one that matters most. A residue panel covers a defined list of analytes. If the panel is thirty compounds and your destination regulates two hundred, "no residues detected" means "none of the thirty we looked for". Ask for the analyte list, not just the conclusion.

Limit of quantification, and why "not detected" is a range

Every analytical method has a limit below which it cannot reliably measure. "Not detected" means "below our limit of quantification", and if that limit is higher than your destination's maximum residue level, a clean report is compatible with a non-compliant consignment.

This is not a hypothetical. It is the single most common way a buyer holds a genuine, honestly issued certificate for goods that are subsequently detained. Specify that the laboratory's LOQ must be at or below the applicable MRL for each analyte, and ask for it in writing on the report.

Onboarding a second source before you need one

The most expensive verification failure is not choosing a bad supplier. It is having exactly one good one.

Single-sourcing a commodity from a single Indian processor concentrates every risk you have just spent four layers reducing: a bad harvest in their catchment, a fire, a labour dispute, a food-safety suspension, or simply their decision that a larger customer matters more than you. None of those are discoverable in advance, and all of them arrive at the worst possible moment.

Qualifying a second source costs the same four layers, and buyers reliably defer it because the first supplier is working. The discipline that makes it happen is to run the second source at a small, standing share — ten or fifteen percent of volume — rather than holding them as a dormant contingency. A dormant second source is a phone number. An active one is a supplier who knows your specification, has shipped against it, and can scale within a season.

The secondary benefit is price discovery. Two suppliers quoting the same specification tell you what the specification actually costs. One supplier tells you what one supplier charges.

Building a supplier file that compounds

Verification done once is a cost. Verification recorded properly is an asset, because the second order is far cheaper to place than the first.

A supplier file worth keeping holds: the verified registrations with the dates you checked them, the factory walk recording or notes, the commercial CoAs you were shown at onboarding, every subsequent shipment's origin and arrival results, the sampling plans used, the contract terms agreed, and a short note after each shipment on what went well or badly.

Three shipments in, that file answers the question every buyer eventually asks — *is this supplier actually good, or have we just been lucky?* — with evidence rather than impression. It also survives the departure of whoever built the relationship, which is the point at which most sourcing knowledge in a small company quietly evaporates.

Where Katalyzon fits

Everything above is doable from outside India. It is also slow, and it is hard to do well in a language and a business culture you do not work in daily.

That is the part we do. Our verification tiers run S0 to S4 — identity and registration first, then capability and capacity, then physical and transaction checks — and every record carries its source, the date it was accessed and a confidence rating. You can see the evidence, including its gaps, on any supplier storefront.

If you would rather hand over the whole process, post a requirement with the product, volume and destination. Our desk reads it, matches suppliers who can actually serve it, and coordinates sampling and third-party inspection. Your details stay masked until you accept a quote.

For how the tiers are defined and what evidence sits behind each, see our methodology and our note on what a supplier must show to earn each tier.