How to write treated vs untreated so a plant trusts the result

A test report is only as useful as the trust it creates. For a corporate buyer, trust comes from method, lab, control, and quantification. Everything else is commentary.

Start with the method and standard

Name the standard. ASTM, ISO, EN, or the buyer's internal method. If you used a modified version, say what you changed and why.

Bad: "We tested water resistance."

Good: "Water absorption per ASTM C642, 28-day immersion, on treated and untreated 4x4x16 inch prisms from the same batch."

The standard is what lets their quality team reproduce or challenge the number.

Third-party or independent verification

If the test was done in your lab, say so and note the equipment calibration date. Better: an independent lab or the buyer's own lab ran the measurements.

When you cannot get third-party for the pilot, at minimum have the buyer witness the testing or split samples. Document the chain of custody.

Control versus treated, always side by side

Never report only the treated result. A buyer needs to see the delta.

Table format that works:

| Condition | Water absorption % | Method | n= |

|-----------|--------------------|--------|----|

| Untreated control | 4.8 | ASTM C642 | 5 |

| Treated (our sealer) | 1.9 | ASTM C642 | 5 |

| Delta | -60% | - | - |

Include standard deviation or range if you have it. A single number without variation looks suspicious.

Quantify, do not qualify

Avoid: "dramatically improved," "excellent performance," "significantly better."

Use: "reduced absorption by 2.9 percentage points (60% relative reduction)."

If the result is within noise of the control, say that plainly. A clean no is more valuable than a soft yes.

Include the conditions that matter

• Substrate or mix design
• Application rate and method
• Cure time and temperature
• Age at test
• Any anomalies noted during the run

A buyer who sees the conditions can judge whether the result applies to their line.

The one-page summary plus the full data

For the buyer and for investors, produce two versions:

1. 1A one-page summary with the headline number, the table, the method, and the date.
2. 2The full lab report or data package as an appendix.

Send the summary first. Keep the full package ready.

What kills trust

• Results with no method named.
• No control data.
• "We tested it and it worked great."
• Charts with no axis labels or sample size.
• Dates that are six months old with no re-test plan.

A buyer will assume the best result was the one you chose to show. Give them the full picture so they do not have to assume.

The report is not marketing. It is evidence. Write it so a skeptical engineer can defend the number to their boss.