The quote said 98.5%. The certificate that arrived nine days later said 96.1%, and it covered a batch that had already gone into a dose-response series. We repeated the work. The reagent cost was trivial; the three weeks were not, and the conversation with a programme lead who had built a timeline around that data was worse than either. That gap between a purity figure quoted at the point of sale and a purity figure backed by an analytical record is something procurement has been absorbing quietly for years. Most of the time it costs nothing. Nobody checks, the compound performs, the work moves on. When it does cost something, the cost lands where no purchasing report tracks it: a repeated assay, a missed milestone, a dataset a reviewer declines to accept. We book the reagent spend and never book the consequence, which is a large part of why the problem has survived this long.
Somewhere along the way “COA available on request” became an acceptable answer, and I have never followed the logic. The certificate already exists. It was generated when the batch was released, months before I went looking for a supplier. Holding it back until money changes hands is not a technical limitation, it is a commercial preference, and it asks a buyer to commit a budget against a specification they have not been permitted to read. We would not tolerate that from a contract manufacturer. We would not tolerate it from an instrument vendor. In research reagents it became normal, mostly because unit prices are low enough that nobody thought it worth the argument.
The questionnaires have changed. Five years ago ours covered purity, storage, lead time and little else. What circulates now asks for batch-specific documentation rather than a representative certificate for the product line. The HPLC trace, not the summary figure derived from it. Mass spec confirmation of identity. The name of the laboratory that ran the analysis, and whether that laboratory is the supplier’s own or genuinely independent. Retention terms, too: can I pull the certificate for a batch bought fourteen months ago, when a reviewer queries the numbers and the person who ordered it has left. Several institutions I work with have written pre-purchase certificate access into their sourcing criteria outright, which means a supplier who cannot produce one is filtered out before anyone opens the price list. That is a meaningful shift. It moves documentation from a thing you chase after the invoice to a thing that decides whether an invoice happens.
Reproducibility is what pushed this from a procurement irritation to something senior people will sit through a meeting about. Batch-to-batch variation in a synthetic peptide is real, expected and entirely manageable. Undisclosed batch-to-batch variation is a different animal. When a compound behaves one way in the pilot and another way in the confirmatory run, and the only paperwork covering both batches is a generic product specification, the experiment cannot be interrogated. The variable is invisible. It gets written up as biological noise, or quietly dropped, and a small documentation failure becomes a permanent defect in the record. Multiply that across a field and some fraction of the reproducibility problem everyone worries about sitting in a supplier’s filing cabinet, not in anyone’s pipetting technique.
A few suppliers have worked out that this is a competitive position rather than a compliance chore. Pure Peptides lists its batch certificates on the front page of its own site, each one a direct download tagged with the batch number and the measured purity, ahead of any account or any order. The individual company is not the point. The point is that the operational excuse has evaporated. If a supplier of that size can publish batch records openly, a larger one declining to do so is making a choice, and we are entitled to read that choice as information about how the batches themselves are managed.
If you are reviewing suppliers this quarter, skip the question about whether certificates are available. Everyone says yes to that one. Ask instead to see the certificate for the batch currently sitting in their stock, today, before you raise a purchase order. Ask who ran the assay and whether they were paid by the supplier. Ask what happens to that record in two years when somebody comes back to the data. The suppliers are worth keeping answers inside an hour, usually with a link. The rest will tell you it is available on request.
