Before You Make the Claim, Build the Evidence Register

A company releases a promising result. The technical team describes what the experiment demonstrated. The commercial team turns it into a benefit. The pitch team compresses the benefit into six words. By the time the claim reaches the homepage, it sounds broader than the evidence that produced it.

Nobody set out to mislead. Each hand-off removed a condition, and the conditions are what a careful buyer or investor needs to know.

Keep one accessible record of what each public claim means, where the evidence lives and which conditions travel with it. The writing can then stay clear while the conditions stay attached.

Our guide to turning technical evidence into an impact story keeps an Impact Log for impact claims, and our list of ten website claims that need evidence beside them covers the words that drift most often. The register below covers every commercially important claim, from performance to readiness and commercial status. It adds the sign-off, the review dates and the places each claim may appear.

Seven fields for every claim

Field	What it records
Exact proposed claim	The words a customer or investor will see
Status	Tested, independently validated, contracted, forecast, targeted or another category you define
Source	The experiment, contract, report, calculation or approved statement behind it
Conditions	What was measured, where, when and under which circumstances
Boundary	What the evidence leaves unestablished
Owner and date	Who can validate the claim, and when it was last checked
Permitted uses	The channels and audiences the claim is approved for

The status labels are a working start, with no industry certification behind them. Measured energy use, clinical efficacy, product safety and the environmental footprint of a supply chain each need different review. The owner chooses the rules that fit the category and the jurisdiction.

The last two fields turn a list of claims into something a team can run. Permitted uses can be a short code for each channel, such as website, investor deck, sales sheet, press, speech or product label. A claim cleared for a data room under a non-disclosure agreement may be unfit for the homepage, and the register records that.

The conditions field carries the most weight once a claim is shortened. Hannes Junginger, co-founder and chief executive of Carbonfuture, which builds measurement, reporting and verification records for carbon removal, described his company's version of this field in a December 2023 interview with us. “Instead of simply stating that we remove a ton of carbon, we specify the amount removed and how long it will be sequestered.” The duration is the condition. A shorter claim that drops it describes a different result.

Keep four distinctions intact

Where the Evidence Sits Matters gives five commercial labels, and the impact story guide gives six evidence states for results. Use the same vocabulary in the status field, so the register and the page agree.

Put the condition beside the benefit

Place the relevant condition close to the benefit. “Validated under specified pilot conditions” tells the reader more than “fully proven” when only pilot evidence exists. If the uncertainty matters to the decision being asked for, link the method or a technical note.

A claim still under investigation may belong off the public website for now. Roadmaps, target performance and intended markets can appear as ambitions, with the tense and the status explicit.

Review the whole sentence

Scientific review checks the number. A technically correct claim can still mislead through an unfair comparison, a headline broader than the study or the image beside it. The reviewer needs the whole sentence, its placement and the neighbouring visual.

Regulated products also need legal and specialist regulatory review. That authority sits with qualified experts, and a communications studio works alongside them.

Set the triggers for a recheck

A register goes stale when nobody reopens it. Recheck a claim whenever one of these happens.

Cut every version from the register

One trusted record can support the product page, investor deck, sample sheet, partner proposal and conference speech, each at its own length. Different readers need different depth and the same facts. Brighter Future's Story System writes each company story once in depth and then cuts it for the room. Apply the same rule to evidence, and cut every short claim from the verified record.

Our work with DNAir, which captures DNA from the air, shows the rule in a new category. Claiming only what the data supports was one of five decisions that shaped its site, and the founders kept every claim inside the evidence when they carried the positioning into their competition pitch.

The register matters most where a market has no obvious reference class. When a procurement director is asked to approve a novel ingredient, or an investor to back a first plant, the open conditions belong inside the decision.

With the register in place, the writing gets easier. The company can state a clear proposition, show the evidence where it matters and make the remaining uncertainty explicit.

When a team wants its claims checked against the source material before a launch, our science communication work works from that material and flags the claims that need stronger evidence, attribution or clearer boundaries. Scientific, regulatory and legal validation stays with qualified experts.

Questions people ask

Should we publish our methodology alongside marketing claims?

For material claims, usually yes, at least as a useful summary. Show how the number was obtained, which boundary applies and where deeper review is possible. Sensitive intellectual property can stay protected. Publish enough method for the reader to understand what the claim means.

How should we communicate results from an independent validation study?

Name the independent organisation, the scope of its work, the standard or method it used and the result. Independence in one test leaves the rest of the company and product outside that validation. Reproduce only what the validation permits, and link the source where publication is authorised.

How should we explain a life-cycle assessment to buyers?

Explain the functional unit, the system boundary, the data sources, the allocation choices and the comparison. Show which results are measured and which are modelled, with the sensitivity where it matters. The buyer should understand the practical implication and read the assessment as a result for that boundary and comparison.

When should a biotech company rebrand after a scientific breakthrough?

When the breakthrough changes the market position, the first offer, the audience or the credible future of the business. First check how the new result affects applications and commercial readiness. A visual identity changes when the definition of the company has moved.

Sources and further reading

• How to Turn Technical Evidence Into an Impact Story Without Overclaiming, Brighter Future, 2026
• 10 Climate-Tech Website Claims That Need Evidence Beside Them, Brighter Future, 2026
• Where the Evidence Sits Matters, Brighter Future, 2026
• Hannes Junginger, Carbonfuture, founder interview, Brighter Future, 2023
• DNAir case study, Brighter Future
• The Story System, Brighter Future
• Science communication for climate, biotech and frontier technology, Brighter Future