Corrections Policy
CommodityNode publishes research-only commodity market intelligence for supply-chain risk, procurement planning, company exposure mapping, and scenario review. Because commodity data changes quickly and public-source datasets can be revised, we maintain a visible correction workflow for factual errors, stale-source issues, and material methodology clarifications.
What qualifies as a correction?
Material factual errors
Incorrect commodity price references, company names, ticker mappings, units, date labels, source citations, or described market events.
Data freshness issues
Stale snapshots, delayed feed labels, proxy-price notes that are missing, or dashboards that do not make the update cadence clear.
Methodology clarity
Ambiguous wording around model ranges, benchmark sources, correlation windows, scenario assumptions, or limitations.
Policy/safety wording
Any copy that could be mistaken for personalized advice, trading instructions, guaranteed outcomes, or brokerage functionality.
Review workflow and SLA
| Issue class | Target acknowledgement | Target action |
|---|---|---|
| Material factual error on a public page | 1–2 business days | Patch or clearly mark the page after verification; urgent if the error changes the meaning of the memo. |
| Stale or missing data-freshness label | 2–3 business days | Add/update timestamp, proxy-source label, archive wording, or data-limit note. |
| Methodology clarification | 3–5 business days | Update the page, methodology, model-limitations, or affected chart note when clarification is warranted. |
| Non-material typo or style issue | Best effort | Batch with the next editorial maintenance pass. |
How updates are displayed
When a correction materially changes a reader’s understanding, CommodityNode aims to update the affected page with a revised timestamp, a short explanation of what changed, and links to the relevant methodology or limitation page where appropriate. Minor copy edits may be corrected silently when they do not affect data, interpretation, or reader safety.
For model or forecast surfaces, corrections may include revised source labels, a model-version note, an uncertainty caveat, a disabled stale-data badge, or a clearer distinction between direct futures data and proxy benchmark context. CommodityNode does not retroactively present a revised model output as if it had been available before the correction.
What CommodityNode does not do
- We do not correct pages to satisfy a requested market view, trade thesis, or preferred directional outcome.
- We do not provide personalized investment advice, trading signals, brokerage, custody, order execution, or portfolio management.
- We do not guarantee that a correction request will result in a page change; evidence and source quality determine the action.
- We do not remove transparent limitations, uncertainty ranges, or archive labels merely because they reduce marketing polish.
Escalation path
If a correction concerns privacy, account data, payment records, or a legal/policy issue, contact contact@commoditynode.com and include “Policy review” in the subject line. For research methodology disputes, include the source, date, unit, and proposed wording. We prioritize corrections that affect public trust, data accuracy, user safety, and Google/Search publisher-quality surfaces.
Related pages: Editorial Process, Editorial Team, Data & Methodology, Model Limitations, and Risk Disclosure.
Evidence we prefer
The fastest correction reviews include primary or high-quality secondary evidence. Useful evidence includes exchange notices, official agency releases, company filings, investor-relations statements, regulator publications, timestamped market-data screenshots, or a clearly identified methodology document. A social post, unsourced chart, or copied summary can still be useful as a lead, but it normally cannot support a material correction by itself.
Versioning and audit trail
CommodityNode keeps correction handling practical rather than performative. For visible pages, the most important audit trail is the updated public page, the changed timestamp when material, and a clear statement when a previous fact was materially wrong. Internal implementation commits, test artifacts, and deploy evidence may also be used to verify that a correction actually reached production. We do not promise to publish private reporter details, sensitive account data, or raw provider logs.
Priority order
Correction priority is based on user impact. Highest priority goes to errors that could mislead a reader about a source, date, unit, company exposure, model limitation, or research-only boundary. Lower priority goes to wording preferences, stylistic disagreements, or requests to make a page more favorable to a specific thesis. This policy exists to protect readers, not to defend a market view.