What can side-by-side deformulation show about a good batch and a bad batch?
Turn the performance labels into a comparison question
A good batch and a bad batch are starting descriptions of performance, not chemical diagnoses of what changed. The same comparison can concern an accepted and rejected production batch, an old and new supplier, or an earlier and later retained batch of the same intended formulation. We ask what observation created the good or bad label, what material state is being compared and which difference would change your next decision. For example, a manufacturer who wants to investigate sticky new parts can ask whether the suspect compound contains a different plasticizer or processing-aid pattern from an unused accepted reference.
Compare matching material states before comparing peaks
Two batches should represent the same supply-chain stage where possible: pellet against pellet, aqueous concentrate against aqueous concentrate, or cured film against a similarly prepared cured film. A fresh liquid against a dried specimen, or a newly produced part against a storage-aged reference, introduces state differences alongside any formulation difference. Preparation, target layer and sampling location should also correspond, since a filler-rich region can differ from a polymer-rich region within one batch. If the comparison concerns incoming material, retained unused samples are more direct than specimens changed by a different application or cure history.
For a finished-part question, describe both preparation histories and keep differences that cannot be aligned explicit in the brief. The wider development and supplier-substitution context is covered in our benchmarking guide; here the target is the difference between the submitted batches.
Use a paired analytical design
For a side-by-side study, we frame both samples as one comparison, using the same selected preparation and methods within the same analytical batch where applicable. This reduces avoidable differences in analytical conditions, but it does not remove unequal sample histories or make every component equally measurable. Depending on the materials, LC-MS/MS and GC-MS can support comparison of accessible organic ingredients or impurities, while ICP-MS can support element comparisons. We look for supported changes in identified compounds, ingredient families, unresolved peak patterns and inorganic contributions rather than assume the bad batch has one missing ingredient.
A difference peak is an observation to interpret, not a compound identity by itself, and stronger identification depends on suitable reference evidence. The distinction between a named compound and an estimated proportion is explained in our identification guide, and cross-report differences in our report-comparison guide.
Read changed signals without turning them into a cause
A stronger additive signal may indicate more of that ingredient, but recovery, matrix effects and measurement response can also change the signal. A new metal-element contribution can suggest a changed inorganic ingredient or contamination, yet its source still needs to be interpreted against the complete material and sampling context. An unreported peak is not proof that the corresponding ingredient is absent, especially where abundance, separation or method coverage limits the finding. We report supported differences in the received samples; a composition difference alone cannot establish the root cause of a performance failure.
Processing, dispersion, moisture and storage can affect behavior alongside composition, so a difference is a lead for further investigation rather than a complete explanation. A chemically similar result also does not prove equal performance, since the measured component map does not capture every relevant physical feature. For a known formulation with a defined release requirement, our quality-control comparison guide explains the different question posed by a specification check.
Keep the packages, labels and batch identities separate
Submit both batch descriptions together and state explicitly in the declaration that the requested work is a comparison, naming the reference and suspect roles. Keep each batch in its own original sealed packaging where applicable, retain its original label and lot number, and use a separate comparison label that does not replace those records. For prepared cured specimens, state the source batch and preparation history for each, and provide photographs linking each specimen to its source description.
Eligibility is reviewed for each batch and actual form; declaring a paired study does not override the receipt limits for either sample. Used material, spent process fluids, waste and scrap are not received from overseas, while unused production specimens can be proposed for declaration review. Unlisted sample types are reviewed case by case after declaration; provide each material's SDS and disclose any liquid or ingredients requiring further review.
How to start
We review overseas industrial samples before acceptance. Use our industrial analysis page and sample-shipping guide to prepare your declaration. Our reports are research-grade analytical reports from a laboratory that is not accredited. They cannot be used for product registration or legal proceedings.
FAQ
Does a difference prove the supplier changed the formula?
A supported batch difference can justify further investigation, but it does not alone prove a deliberate formulation change because preparation, storage and contamination can also alter the material.
Why analyze both batches together?
Analyzing a batch pair with matching preparation and methods in the same analytical batch where applicable reduces avoidable method-condition differences, making the comparison easier to interpret.
Can matching ingredient lists rule out the failure?
Matching ingredient lists cannot rule out performance differences caused by physical structure, processing or components outside the supported analytical scope.
What batch information should the declaration include?
For a batch comparison, name both products, preserve their lot identities, identify reference and suspect samples and describe each physical form and preparation history.