A documented Dieselgate analysis and traceable test for ensuring environmental claims match product operation, real-world evidence and accountable controls.
Short answer: answer: Dieselgate shows that a claim is governable only when marketing can trace it to how the product behaves in ordinary use, not merely to a certificate, laboratory result or approved phrase. For every material environmental claim, link the exact wording to the operating mechanism, test conditions, real-world verification, accountable engineer and expiry trigger; block publication when any link is absent. Compliance approval cannot rescue a claim when the underlying system is designed to behave differently during the test.
The comfortable lesson is “do not lie.” That is morally correct and operationally weak. Large organisations rarely manage claims through one person deciding whether to tell the truth. They manage them through specifications, tests, certificates, product data, agency briefs, legal review and executive incentives. If those records are disconnected, marketing can publish a technically reviewed statement that is false in the customer’s world.
CDM’s position is uncompromising: marketers making performance or sustainability claims must have challenge rights over product evidence, not just responsibility for wording. They need not become emissions engineers, but they must know which qualified owner attests that the evidence covers normal operation and which independent test could disprove it.
The documented claim-to-product breach
In September 2015, the US Environmental Protection Agency issued a notice of violation stating that Volkswagen had installed software defeat devices in certain model-year 2009–2015 2.0-litre diesel vehicles. The software detected standard emissions testing and altered the operation of emissions controls. EPA later said affected 2.0-litre vehicles emitted nitrogen oxides up to 40 times the applicable standard under some normal driving conditions.
The US Department of Justice said Volkswagen marketed the vehicles as “clean diesel” and environmentally friendly while employees involved in the scheme knew the vehicles were designed to evade emissions testing. In 2017 Volkswagen agreed to plead guilty to three felony counts and pay $4.3 billion in criminal and civil penalties. Separate consumer and environmental settlements followed; the FTC later reported more than $9.5 billion returned to buyers and lessees under its redress programme.
Those are findings and resolutions from US regulators, court records and the company’s guilty plea. They cover defined vehicles, dates and jurisdictions. “Every Volkswagen environmental statement was false” would go beyond that record.
The Claim-to-System Consistency Test
The Claim-to-System Consistency Test uses six linked proofs. Marketing may simplify a proposition for customers, but the evidence chain behind it must become more detailed as the claim becomes broader, more absolute or more consequential.
Proof 1: freeze the exact claim
Capture the headline, qualifiers, images, comparison, audience, channel, geography and campaign dates. “Clean,” “low emissions,” “meets the US standard” and “lower impact than the previous model” create different burdens.
Test the net impression, not isolated words. A qualifying footnote cannot reliably correct a dominant environmental image or an unqualified headline. Have an independent reviewer describe what a reasonable buyer would believe before showing the substantiation file.
Proof 2: identify the operating mechanism
Name the physical, software or organisational mechanism that makes the claim true. For emissions, that includes the control system and calibration across expected operation. For recycled content, it includes material sourcing and mass-balance method. For renewable electricity, it includes the contractual instrument and boundary.
The product owner must attest that the mechanism exists in the production product, not only a prototype, and that material updates trigger reapproval. Marketing should never accept “engineering signed off” without a named artefact and scope.
Proof 3: expose the test boundary
Record laboratory protocol, sample, operating conditions, geography, date, uncertainty and exclusions. Then compare them with foreseeable customer use. A claim based on one controlled condition must not imply equivalent performance everywhere.
Dieselgate is the extreme case: the system detected test conditions and behaved differently. The general control is a test-boundary challenge—what changes outside the certification cycle, and how do we know? This does not diminish legitimate standardised testing; it prevents the standard from being misrepresented as the whole operating reality.
Proof 4: seek disconfirming real-world evidence
Predefine an independent field test, complaint signal, telemetry measure or audit capable of disproving the claim. The West Virginia University work commissioned by the International Council on Clean Transportation identified a discrepancy between on-road and laboratory emissions, according to the DOJ chronology.
Governance fails when contrary evidence is treated as a communications problem. Route it to a claim incident owner, pause affected creative when the discrepancy is material, and preserve the evidence. The same rule applies to creator scripts and retailer listings supplied by the brand.
Proof 5: separate owner from challenger
Name the product evidence owner, legal reviewer, marketing publisher and independent challenger. The person rewarded for launch timing should not be the sole judge of conflicting test evidence. Give the challenger direct escalation to a senior decision owner or board-level risk function for high-consequence claims.
An approval tool is only a routing system. It becomes governance when the approvers have relevant competence, access to source evidence and authority to say no.
Proof 6: set expiry and propagation
Every claim approval needs an expiry date and change triggers: software update, supplier change, new model year, standard revision, new jurisdiction, contrary field evidence or regulator inquiry. Maintain a list of every live placement and derivative so a correction can propagate to websites, dealers, ads, creator content and archived sales materials.
An evergreen sustainability page built on expired product evidence is not a documentation problem. It is a live claim problem.
Reader asset: claim-to-operation trace
| Trace field | Required record | Dieselgate warning |
|---|---|---|
| Customer claim | Exact words and net impression | “Clean diesel” carried an environmental promise |
| Mechanism | Production control and responsible owner | Product operation contradicted the promise |
| Test | Protocol, conditions, sample and uncertainty | Test behaviour differed from road behaviour |
| Field check | Independent disconfirmation method | On-road testing exposed the discrepancy |
| Approval | Owner, qualified challenger and decision date | Formal review is weak without system truth |
| Triggers | Expiry, product change and contrary-evidence alert | Claims persisted while the underlying issue was concealed |
| Propagation | Inventory of ads, pages, dealers and creators | Correction requires finding every derivative claim |
Require all seven rows before the campaign receives an asset ID. For a high-consequence environmental or health claim, CDM recommends direct inspection of the primary test report by a qualified specialist and a real-world evidence check. This is operational guidance, not a replacement for the FTC Green Guides, local environmental-claims rules or legal advice.
What marketers should refuse
Refuse a substantiation summary that contains no test conditions. Refuse a certification used outside its scope. Refuse an absolute claim supported only by an average. Refuse an image or creator script that broadens a qualified engineering statement. Refuse an approval whose product version does not match the one being sold.
Most importantly, refuse the idea that marketing is merely the messenger. A claim changes demand and price; that commercial effect makes evidence governance part of marketing operations. The marketer is not expected to detect hidden software unaided. The marketer is expected to build a process where named experts can be challenged, conflicting evidence cannot disappear and live claims can be stopped.
Related guides
Frequently asked questions
What was misleading about Volkswagen’s clean-diesel marketing?
US authorities said Volkswagen marketed affected diesel vehicles as clean and environmentally friendly while software detected emissions tests and changed control behaviour. Under some normal driving conditions, EPA said certain affected vehicles emitted nitrogen oxides up to 40 times the standard. Volkswagen pleaded guilty to criminal charges connected with the scheme, and major civil and consumer settlements followed.
The caveat is scope: the official records identify particular diesel engines, model years and markets. Use those boundaries rather than converting the case into a claim about every Volkswagen vehicle or every environmental statement the company made.
Does regulatory certification prove a marketing claim?
Certification proves only what the named scheme, product, period and test establish. Marketing must understand its scope and avoid turning a limited compliance result into a broader performance or environmental promise. Verify that the certified production version matches the advertised one and seek evidence from foreseeable operating conditions.
Record that comparison explicitly. The caveat is that marketers are not expected to reproduce every specialist test. They need a qualified owner, primary report, independent challenge and change control that make reliance reasonable and traceable.
Who should approve an environmental marketing claim?
Use at least a business owner, technical subject-matter owner, legal or regulatory reviewer and marketing publisher, with independent sustainability review for material claims. Each signs a different proposition: commercial use, system truth, legal sufficiency and faithful communication. One omnibus “approved” status hides those distinctions. Record dissent as well.
The exception is a narrowly factual statement with low consequence, where proportional review may be lighter. Even then, retain the source, scope, owner and expiry so the fact cannot remain live after the product changes.
How often should green claims be revalidated?
Revalidate on every material product, supplier, methodology, software, standard or jurisdiction change, whenever contrary evidence appears, and at a scheduled interval tied to evidence volatility. Annual review is a practical ceiling for many continuing claims, not a universal safe harbour.
A five-year-old lifecycle assessment may remain informative for an unchanged product, while a software-mediated performance claim can expire with one release. Record every renewal decision. The caveat is law: specific programmes and jurisdictions may impose different recordkeeping, update and disclosure obligations.
What should marketing do when new evidence contradicts a live claim?
Pause new distribution, preserve the evidence, identify every live derivative, and convene the technical and legal owners to assess materiality. Correct the source claim first, then propagate the decision to ads, product pages, sales tools, creators and partners. Record the reason and effective time.
Do not quietly narrow language while leaving old claims in paid media. Verify removal channel by channel. The caveat is safety or regulator involvement, where counsel may control communication and preservation. Operational containment should still begin immediately within that direction.
Are broad words such as “clean” or “green” ever safe?
They are high-risk because reasonable buyers may interpret them across several environmental dimensions and over the full product lifecycle. A specific, qualified claim—naming the attribute, comparison, boundary and evidence—is usually more governable. Whether a term is lawful depends on context and jurisdiction; the FTC Green Guides and other authorities provide relevant rules.
The caveat is that specificity alone does not cure weak evidence. “Forty per cent lower” is more precise than “green,” but it is more misleading if the baseline, method or operating condition is wrong.
Next decision: How Do You Correct One False Claim Across Website, Email, Social and Creator Content?
Related reading: How Do You Build a Claim Evidence Register for Marketing Content? · How Do You Check Whether a Marketing Statistic Is Safe to Publish? · Why Do Marketing Campaigns Fail Even When the Strategy Looks Sound?
Sources and research notes
- EPA notice of violation, 18 September 2015 — primary regulatory notice describing defeat devices; checked 26 September 2026.
- EPA overview of Volkswagen violations — affected US vehicles, chronology and emissions description; checked 26 September 2026.
- US Department of Justice plea and penalty announcement — charges, plea terms, marketing conduct and chronology; checked 26 September 2026.
- Volkswagen plea agreement and statement of facts — primary court agreement, including clean-diesel marketing facts; checked 26 September 2026.
- FTC final redress summary — consumer-marketing case and reported repayment; checked 26 September 2026.
- FTC Green Guides — US environmental-marketing guidance hub; checked 26 September 2026.
- Limitations: This article focuses on documented US proceedings and is not a complete global Dieselgate history. Claim rules vary by jurisdiction; the Consistency Test is CDM operational guidance and requires legal and technical review.
This article is editorial guidance. Apply the principles in proportion to your market, evidence, and responsibilities.
