What should a chemical product recommendation to a buyer include?
Summary
A product recommendation to an industrial customer needs 5 sections: problem and requirements, recommended product with evidence, rejected candidates, application conditions, next steps. The structure holds when a supplier or distributor proposes a chemical product for one named process, and the customer must decide on a trial, an approval or a rejection from the document itself.
NASA decision analysis guidance says an option that does not meet mandatory criteria should be disregarded, and that options which drop out should still be documented as considered (NASA Systems Engineering Handbook, section 6.8, accessed 2026).
Kepner-Tregoe decision analysis splits objectives into MUSTs, which screen alternatives out, and WANTs, which compare the alternatives that remain (Kepner-Tregoe, 2021).
US OSHA requires chemical manufacturers and importers to provide distributors and employers a safety data sheet with the initial shipment of a hazardous chemical (US OSHA, 29 CFR 1910.1200(g)(6)(i)).
Under EU REACH, a safety data sheet shall be provided free of charge no later than the date on which a covered substance or mixture is first supplied (Regulation (EC) No 1907/2006, Article 31(8)).
Who reads a chemical product recommendation, and what do they decide?
A chemical product recommendation is usually read by three people at the customer, and each makes a different decision from it. The process or maintenance engineer who owns the problem decides whether the product addresses it under the plant's own conditions. The quality, safety or environmental reviewer decides whether the safety data sheet, approvals and site rules allow the product onto the site. The buyer decides whether to open a trial order and on what terms.
NASA's decision analysis guidance makes the same point for any recommendation: the specific decision-making body "should also be well defined" before the approach to the decision is chosen (NASA Systems Engineering Handbook, section 6.8). For a chemical product recommendation, name these readers in the first section and state what each one is asked to approve.
The readers also judge the recommendation without the supplier in the room. A line such as "proven in similar plants" with nothing attached gives the reviewer nothing to check, so the decision waits for another meeting. Write each claim so that a reviewer who missed the sales call can trace it to an attached document, a stated test method or a customer record. The recommendation succeeds when the customer can approve a trial, ask a specific question or reject the product from the document alone, and fails when the next step is a general request for more information.
Which sections does a chemical product recommendation need?
A chemical product recommendation needs five sections in a fixed order, and each section carries its own evidence. The table below shows what the customer decides from each section of the recommendation and what to attach so that the decision can be made.
Section | What the customer decides | Evidence to attach |
|---|---|---|
1. Problem and requirements | Whether the supplier understood the process and its limits | Customer specification, equipment maker limits, site rules, measured conditions with dates |
2. Recommended product and evidence | Whether the product meets every must-have requirement | Technical data sheet, safety data sheet, test reports with method and conditions, approvals |
3. Rejected candidates | Whether the search was wide enough and the ranking fair | Candidate table with the requirement each one fails and the evidence for it |
4. Application conditions | Whether the plant can run the product as specified | Dilution, dose, temperature range, compatibility notes, handling and storage from the safety data sheet |
5. Next steps | What to approve now and what the trial must confirm | Open checks, trial scope, owners, documents still to come |
The section most easily left out of a chemical product recommendation is the third, rejected candidates. Without it, the reviewer cannot tell whether the recommended product won a comparison or was the only one considered. Keep sections 1 and 2 apart as well: the requirements are the customer's, written before any product is named, and the evidence in section 2 is checked against them line by line.
Sections 4 and 5 carry the practical weight of a chemical product recommendation. Application conditions tell the plant whether it can run the product with its own water, equipment and staff, so take them from the supplier's documents and the safety data sheet rather than from a general description. Next steps turn the recommendation into a decision: name what the customer is asked to approve now, which checks the trial must close, who owns each one and which documents are still to come. Fix the order of the five sections before writing any of them, and keep each one short enough that a reviewer can check it in one reading.
How do you separate must-have requirements from preferences in a product recommendation?
You separate must-have requirements from preferences in a product recommendation by asking one question of each line: if a candidate fails this, is it out, whatever else it offers? A yes makes the line a must-have; a no makes it a preference that can only rank the candidates that pass.
Kepner-Tregoe decision analysis uses the same split. Its MUST objectives are "mandatory, minimum standards", and its WANT objectives "provide a comparative picture of alternatives"; in the firm's words, "the MUSTs decide who gets to play, but the WANTs decide who wins" (Kepner-Tregoe, 2021). NASA's guidance treats decision criteria as either mandatory or enhancing and states: "An option that does not meet mandatory criteria should be disregarded." It adds that "Criteria should be objective and measurable" (NASA Systems Engineering Handbook, section 6.8).
Write each must-have requirement with a value, a unit, a test method and the source of the limit, such as the customer specification, the equipment maker's manual or a site permit. A must-have that cannot be measured is a preference in disguise. Kepner-Tregoe also advises checking whether a MUST is truly valid before including it (Kepner-Tregoe, 2021), because an inflated must-have removes good candidates for no reason.
Then test each preference against the ranking. If no realistic value of a preference would change which candidate ranks first, record it for the customer but leave it out of the scoring, so the reader sees which factors decided the outcome.
How should a product recommendation record why candidates were rejected?
A product recommendation should record each rejected candidate in one row: the candidate, the must-have requirement it fails, and the evidence that shows the failure. NASA's decision analysis guidance gives the reason: "Some options may quickly drop out of consideration as the analysis is conducted. It is important to document the fact that these options were considered" (NASA Systems Engineering Handbook, section 6.8).
The candidate table below decides which candidates are out and why; the candidates, limits and values are illustrative.
Candidate | Requirement it fails | Evidence |
|---|---|---|
Candidate A, solvent-based (illustrative) | Flash point at or above 60 °C, site storage rule (illustrative) | Safety data sheet section 9 gives a lower flash point (illustrative) |
Candidate B, water-based (illustrative) | Compatible with the yellow-metal parts on the line (illustrative) | Supplier test report shows staining on copper alloy (illustrative) |
Candidate C, water-based (illustrative) | None failed; ranked second on drying time, a preference (illustrative) | Technical data sheet typical value, not a specification (illustrative) |
The key row in this illustrative table is Candidate B: it is out on a must-have, compatibility with yellow metal, and the evidence is a named test report the reviewer can request. Record a candidate as failing only when the evidence shows it; a missing value is an open check, not a failure. When a lower-ranked candidate is recommended, NASA's guidance says "an explanation should be provided as to why the lower score is preferred" (NASA Systems Engineering Handbook, section 6.8). Put that explanation in the same section. Keep the rejected candidates in the final document even when the customer only asked for one product, because the same rows answer the reviewer's first question and save the next search from starting again.
Which checks should a product recommendation leave for the trial?
A product recommendation should leave for the trial only the checks that the documents and lab work could not settle and that could change the decision. Everything the data sheets, test reports and customer records already answer belongs in section 2, not in the trial.
Lab and data-sheet evidence has known gaps. Nexdigm notes that a product that performs well under laboratory conditions "may behave differently at industrial scale", and lists viscosity, curing behaviour, mixing, shelf life, phase stability, equipment compatibility and processing speed as factors that can affect qualification (Nexdigm, accessed 2026). Data-sheet values are a second gap. One maker's product data sheet states that typical properties "do not constitute a specification" (ExxonMobil, Mobil Xtra EP WT product data sheet, accessed 2026), and another tells specification writers that its values "are not intended for use in preparing specifications" (Dow, DOWSIL 795 product data sheet, accessed 2026). A must-have met only by a typical value is an open check.
NASA's guidance gives the test for which checks matter: report whether "reducing them could credibly change the ranking of the alternatives" (NASA Systems Engineering Handbook, section 6.8). For each open check, write what is unknown, why the documents cannot settle it, the trial measurement, and the result that would reverse the recommendation. The Australian competency unit for managing trials sets trial metrics the same way, starting with determining the information required from the trial (training.gov.au PMC557001, 2016). Keep the list to the few checks that could reverse the choice.
What does an illustrative chemical product recommendation look like?
An illustrative chemical product recommendation for a rust preventive used between machining steps fits on two pages and follows the five sections in order. Every product, value and date in this example is illustrative and comes from no real case.
The table below shows the content of each section in the illustrative recommendation.
Section | Content (illustrative) |
|---|---|
1. Problem and requirements | Steel parts show flash rust within 48 hours between machining and assembly. Must-haves: no rust on parts for 7 days indoors, compatible with the cutting fluid in use, flash point at or above 60 °C (site rule). Preference: drying time under 30 minutes (illustrative) |
2. Recommended product and evidence | Water-based product X. Safety data sheet, technical data sheet, humidity cabinet report with method and duration, cutting fluid compatibility note from the supplier lab (illustrative) |
3. Rejected candidates | Candidate A out on flash point; candidate B out on copper alloy staining; candidate C passed all must-haves and ranked second on drying time (illustrative) |
4. Application conditions | Dip at 5 to 10 percent in soft water, 20 to 40 °C, change bath on a measured concentration limit; storage and handling per the safety data sheet (illustrative) |
5. Next steps | Trial on one line for 2 weeks. Open checks: protection on the plant's own parts in its humidity, and residue at assembly. Owners: process engineer and quality (illustrative) |
The key row of this illustrative recommendation is section 5: it names two open checks, rust protection on the plant's own parts and residue at assembly, because the humidity cabinet report could not answer either for this plant. Drying time stays out of the trial checks: it is the preference that ranks X above candidate C, and candidate C stays named as the fallback, so a slower drying time on the line would not change the trial decision (illustrative).
Read the illustrative recommendation the way each customer reader would. The process engineer checks that the three must-haves in section 1 match the line, and that the humidity cabinet report in section 2 states its method and duration. The safety reviewer finds the flash point rule in section 1, the reason candidate A was rejected in section 3, and the safety data sheet attached in section 2. The buyer sees a trial of fixed length with named owners in section 5. None of them needs a follow-up call to approve or reject the trial, which is the test the five sections are written to pass.
Why are chemical product recommendations rejected?
Chemical product recommendations are often rejected for reasons a reviewer can see on the page, even when the product itself would work. Six patterns are worth checking for, and each one leaves a reader at the customer unable to approve.
Requirements copied from the supplier's catalog. If section 1 restates the product's strengths instead of the customer's limits, the engineer cannot check that the problem was understood.
Typical values presented as guarantees. Makers state that typical properties do not constitute a specification (ExxonMobil, Mobil Xtra EP WT product data sheet, accessed 2026), so a reviewer who notices the difference treats the claim as unproven.
No rejected candidates. A recommendation that shows one product invites the question of what else was considered.
Missing safety data sheet or regulatory status. In the United States the safety data sheet must be provided with the initial shipment of a hazardous chemical (US OSHA, 29 CFR 1910.1200(g)(6)(i)), and in the EU it is due no later than first supply where REACH Article 31 applies (Regulation (EC) No 1907/2006, Article 31(8)). Reviewers often ask for it before a trial.
Preferences scored as must-haves, which removes candidates without a reason the customer accepts.
No open checks. Without a short list of what the trial must confirm, the trial scope grows and the decision date moves.
Each fix is a section of the recommendation, not more text.
Chemical product recommendation checklist
Run this chemical product recommendation checklist before the document goes to the customer. Each line should be true of the final version.
The readers are named, with the decision each one is asked to make.
The problem is stated in the customer's terms, with measured conditions and their dates.
Every must-have requirement has a value, a unit, a test method and the source of the limit.
Preferences are listed apart from must-haves, and only preferences that could change the ranking are scored.
The recommended product's evidence is checked against each must-have, line by line, with the document named.
Values taken from a technical data sheet are marked as typical or as specification.
Every rejected candidate has the requirement it fails and the evidence; a missing value is listed as an open check.
The application conditions come from the supplier's documents and the safety data sheet, not from memory.
The safety data sheet and any approvals the site asks for are attached or dated as to follow.
Each open check names what is unknown, the trial measurement and the result that would reverse the recommendation.
Next steps name owners and the next decision.
Lubinpla's product recommendation keeps rejected candidates with their reasons and registers a product proposal PDF, with the ranking and the trial check items, to that customer's company record for review.