You recommend from documented product data and say plainly what has not been tested. A defensible recommendation rests on the grade's specification, its documented application data and its known compatibility constraints, with the untested conditions named for the customer to validate. Kimia, a chemical intelligence platform for chemical suppliers and distributors, keeps each answer attached to the source documents behind the grade.
An untested application still has documented evidence
A question about an application no one has run is rarely a question with no evidence. The grade still carries a specification, a documented chemistry, application data from adjacent uses and compatibility constraints recorded across its history. A recommendation built on those is an inference from documented properties. That inference is legitimate technical selling when it is labelled as an inference. It becomes a liability the moment it is dressed up as tested fact.
In practice: a customer formulating a waterborne wood coating asks about a polyurethane dispersion the supplier has only validated in leather finishing. Film formation, hardness development and solvent tolerance are all documented in the specification. The wood-coating application is not. Both facts belong in the answer.
Every statement traces to a document
A recommendation for an untested formulation holds up only when each claim in it points at a source: the TDS, an application guide, a compatibility table, a lab report from an adjacent project. Where the documentation is silent, the silence is information. What the specification does not state is precisely the list of things the customer's own laboratory has to establish.
In practice: a rep proposing photoinitiators for a customer's new UV-curable material cites absorption range, recommended loading and cure response from the specification. Migration behaviour under EU food-contact rules is absent from the file, so it goes to the customer flagged as untested.
A named gap strengthens the recommendation
The honest answer has three parts: the grade being proposed, the documented basis for proposing it, and the conditions nobody has tested. State all three and the gap becomes a scoped validation step. Hide the third and the customer discovers it mid-trial, where a wrong recommendation fails under process conditions or regulations and the account remembers it. Buyers validate every new material in their own formulation anyway. What they judge is whether your claims survived.
In practice: an adhesion promoter for polypropylene carries peel-strength data measured at 23°C. The customer's line runs at 60°C. The recommendation states the measured basis, flags the temperature difference and proposes the customer's trial as the test that closes it.
Finding the evidence is the slow part
The documented evidence usually exists. It sits across TDS revisions, application guides, lab reports, regulatory files and old email threads, and no rep holds it in memory across hundreds of grades. So an untested-formulation question either queues behind a technical expert for days or gets answered from recall. Both routes lose. Buyers comparing suppliers progress with whichever one answers accurately first.
In practice: a customer asks which other formulations in the range contain a raw material they already stock, a specific plasticiser. The answer exists in dozens of separate TDS files. A rep who has to open each one loses the conversation before finishing the third document.
Kimia keeps the source attached to the answer
Kimia connects product, regulatory and formulation knowledge so commercial teams answer customer questions without waiting on technical experts. Every response is grounded in your own knowledge base with source attribution, so a rep can trace a recommendation to the document behind it. Confidence scoring flags when the system is less certain, and the team knows when to verify. A general model like Claude or Copilot starts cold on your chemistry. Kimia doesn't.
In practice: a rep asks the Technical Sales Assistant which grades fit a customer's new application. Each candidate returns with the specification and application guide it was drawn from, and when a decisive attribute is missing from the data, Kimia asks for it instead of guessing.
What changes for the sales team
Defensible answers: every recommendation traces to a named source document.
Visible gaps: untested conditions are stated before the customer finds them.
Faster first response: documented evidence surfaces during the conversation rather than days later.
Scoped trials: sample requests arrive with the open question already defined.
Why the first accurate answer wins
391%: conversion lift from responding within 1 minute, per research from Velocify.
78%: share of buyers who choose the first company to respond, per Velocify.
Who runs Kimia
Kimia is already live with enterprise customers including Bostik, Univar Solutions and Stahl. Univar Solutions deploys Kimia to accelerate technical sales. At Bostik, Aldric Tourres, Global Director of Digital, says: "Kimia has become our partner of choice within Bostik for scaling technical expertise globally."
A recommendation for an untested formulation is a judgement built on documented evidence, and it reads best when it looks like one: this grade, this data, this gap, this trial.
FAQ
Can you recommend a product for a formulation that has never been tested?
Yes, when the recommendation is grounded in documented product data and the untested conditions are stated openly. The grade's specification, application data and compatibility constraints support an informed proposal. The customer's own validation trial then covers the conditions no documentation reaches.
What should a recommendation for an untested application include?
Three things: the grade being proposed, the documented basis for proposing it, and the specific conditions the supplier has not tested. Each claim should trace to a source such as the TDS or an application guide, and the untested conditions define the customer's validation trial.
Does admitting a lack of testing weaken the recommendation?
No. Buyers validate every new material in their own formulation before adopting it, so the trial happens whether the gap is named or hidden. A stated gap positions the supplier as the party who understands the application. A hidden gap surfaces mid-trial and costs the account's trust.
How does Kimia support recommendations for untested formulations?
Kimia grounds every response in the company's own product, regulatory and formulation data, with source attribution so a rep can trace each answer to the document behind it. Confidence scoring flags when the system is less certain, so the team knows when to verify before the recommendation reaches the customer.
