Spotting a bad source

What this guide is not

It is not a ranking, it is not a shortlist, and it will not name a supplier we think is fine, because we do not think that about any of them and we would be lying if we implied otherwise. What it is: a way of noticing that something is wrong, assembled from what members in the sourcing and quality circles have got right and — more usefully — from what they got wrong first.

The framing that has helped people most is to stop asking "is this one good" and start asking "what would this seller have to do for me to catch them out". Almost every reassurance in this market is cheap to fabricate. The few that are expensive to fabricate are the ones worth your attention, and there are not many of them.

Everything here applies to the supplier’s conduct. None of it can tell you what is in a vial. Those are separate questions and conflating them is the commonest mistake we see.

Signals that feel informative and are not

Members have been burned by every item on this list, usually while feeling careful.

• Photographs of a laboratory. Stock imagery, or a real facility belonging to somebody else. A photograph of a clean room is not evidence that your powder passed through one.
• A large review count. On a site the seller controls, this is a design decision. On a site the seller does not control, it is still a self-selected population of people who received a parcel and not of people who tested its contents.
• Company registration details. Registration establishes that a company exists. It says nothing about what it manufactures, whether it manufactures at all, or whether it is a trading front for somebody else.
• Superlative purity figures. A row of near-identical high nines across an entire catalogue is a marketing artefact, not a manufacturing achievement. Real analytical results scatter.
• Corporate name resemblance. Trading under a name that resembles a large listed group is not an association with it. If a relationship cannot be checked, it should carry no weight at all.
• Urgency. Countdown discounts, bundle pricing that expires, "last batch of this lot". Pressure to decide quickly is not a quality signal in any market and is a warning in this one.

Signals that are worth something

Thin list, deliberately. These are the things that cost a dishonest seller real effort.

A specific answer to a specific question. Ask which lot your vials will ship from and ask for the report for that lot. A supplier who can produce lot-level documentation is doing something operationally different from one who re-sends the catalogue certificate. Members report that the second response is far more common, and it is itself the answer.

Consistency across independent surfaces. When published analysis, member reports over a period of years, and results commissioned by people with no relationship to the seller all point the same way, that is weak evidence rather than none. When they disagree, believe the unflattering one.

Willingness to say no. A supplier who will state plainly that nothing has been tested for endotoxin or sterility is telling you the truth about a limitation. That is more informative than one who implies coverage nobody has.

Behaviour when something goes wrong. The parcel that arrived short, the cracked stopper, the seizure — what actually happened next, as reported by more than one member, over more than one incident. This is the only part of a supplier’s conduct you can observe rather than infer.

Document forensics without a chemistry degree

You do not need to read a chromatogram to spot the usual problems on an analysis document. You need to read the boring parts.

Check that the sample identifier on the certificate matches the lot you are being offered, and that both appear somewhere other than the certificate. Check the dates: a report dated before the synthesis date, or a synthesis date years before the shipment, both mean something. Check that the client named on the report is the company you are dealing with, because a report commissioned by a third party for a different batch is a document about somebody else’s material. Check that the method section exists at all — a certificate with a purity number and no stated method, column, wavelength or instrument is a claim, not a result.

Then check the file itself. A photograph of a screen, a re-flattened image with a different font on one line, a page count that does not match, missing page numbers, a signature block with no analyst name. Alteration is usually clumsy. Reading a COA line by line is the long version of this paragraph and it is the guide members say changed how they look at these documents.

What diligence cannot rule out

Here is the part that sellers never write. You can do every check in this guide, get satisfactory answers to all of them, and still be exposed to the things that matter most.

None of it establishes sterility or endotoxin load. None of it establishes that your vial came from the tested lot. None of it establishes chain of custody — who synthesised the material, where it was repackaged, how many hands it passed through, or whether the company you are talking to has ever seen the inside of the building it came from. None of it establishes batch-to-batch consistency, which is precisely what fails when a supplier that was fine for two years quietly changes its own supplier. And none of it gives you a remedy.

Sunil’s formulation, repeated until members can say it back to him: diligence can lower your chance of dealing with an obvious fraud. It cannot make an unregulated injectable into a regulated one. If you find yourself feeling safe at the end of a checklist, the checklist has done something it was not designed to do.

Walking away is a result

The single most useful behaviour we have seen in the sourcing circle is the member who stopped. A woman in Ireland had two orders six months apart behave differently once reconstituted and decided not to spend a year diagnosing it. Several people thought that was the sensible response, and it is worth saying why: she had no analytical capability, no leverage, and no way to escalate, so investigating further would only have produced a more elaborate guess.

You do not owe a supplier the benefit of the doubt. You are not obliged to complete a purchase because you have spent an evening researching it — the research is a sunk cost and the parcel is not. And "I could not get a straight answer about the lot, so I did not order" is a complete and respectable reason, which our circle hears roughly once a fortnight and applauds every time.

If you want the recorded member experience rather than a checklist, the reference pages at vendors set out what has been reported and what has never been established, and the sourcing-safely circle is where the conversation happens.