Editorial policy

Sourcing hierarchy

Every factual claim is tiered, and the tier a claim needs depends on the risk it carries.

Tier 1 — primary government and statute. OSHA, DOT and PHMSA, EPA, CDC, state health departments, state victim compensation boards, the DOJ Office for Victims of Crime, and state statutes.

Tier 2 — standards bodies and industry institutions. IICRC, ANSI, RIA, insurance industry bodies, the National Association of Crime Victim Compensation Boards.

Tier 3 — estimates only, always attributed by name. Cost aggregators and cleanup company publications.

Vendor-authored content is the dominant source in this niche, and we treat all of it as Tier 3. Every cleanup company blog has a commercial reason to overstate insurance coverage and understate your alternatives. We never cite a single vendor source for a coverage claim.

Risk tiers

R3 — legal and regulatory. State licensing, medical waste transport, compensation eligibility, caps and deadlines, disclosure law. Requires a primary government source and carries a dated Jurisdiction Notice naming the agency to verify with.

R2 — safety. Exposure risk, protective equipment, pathogen handling, employer obligations. Requires an OSHA, CDC or standards-body source.

R1 — general. Costs, definitions, comparisons.

Our build refuses to publish an R3 page without a Jurisdiction Notice, and refuses to publish one whose only sources are Tier 3.

The tone mandate

This governs every page touching unattended death, suicide, homicide or hoarding. It is enforced in our build pipeline: a page that breaks it fails the build.

1. No sensational language. No “gore”, “grisly”, “shocking”, “horrifying”. No clickbait detail and no dramatised scenarios. We never describe decomposition in visceral terms — we describe its practical consequences, which is which materials have to be removed and why.
2. No trauma imagery. No before-and-after photographs, no scene photography, no stock images of body bags, police tape or blood. Ever.
3. Person-first, dignity-preserving language. “A loved one who died”, not “the body”. “The person who lived here”, not “the deceased”. “Died by suicide”, never “committed suicide”.
4. Crisis resources on every sensitivity-flagged page, visible without scrolling on a phone: 988 Suicide & Crisis Lifeline, available in Spanish, and Crisis Text Line. Hoarding pages lead with a referral to a licensed mental health professional, framed as the primary intervention rather than as an afterthought to cleaning.
5. Hoarding is a recognised disorder, not a mess. Never “filthy”, “disgusting” or “lazy”, and no before-and-after shaming.
6. Cost content never reads as extraction. Every pricing page links to insurance and victim compensation above its cost tables, never below them.
7. No urgency mechanics. No countdowns, no “act now”, no scarcity framing, no live-chat pop-ups. The reader is not shopping.
8. Sensitivity review is a gate, not a courtesy. Any page carrying the flag requires a named human reviewer and cannot be batch-generated.

Review process

Every page carries a named author and a visible last-reviewed date. Pages that touch death, suicide or hoarding carry a second name: the sensitivity reviewer, who checks the page for harm rather than for accuracy. Those are different jobs and we do not collapse them into one.

R3 pages are re-checked against their primary source on a scheduled cycle, because caps and deadlines change and a page that was right last year can cost someone money this year.

Translation

Our Spanish pages are not machine translations. Euphemism norms around death differ substantially between countries of origin, and a literal rendering can read as callous to one reader and evasive to another.

No page about death, suicide or trauma may be machine-translated here. Those pages are written in Spanish and reviewed by a native speaker briefed on this tone mandate. Where a page exists only in English so far, the Spanish URL says so plainly and links the English original rather than shipping a translation nobody has read.

AI use

Automated tooling is used in this project for corpus processing — deduplicating and clustering search questions, extracting entities, and flagging which pages need sensitivity review. It is also used in drafting, under human direction.

No page publishes on automated output alone. Every R3 claim is checked against a primary source by a person, and every sensitivity-flagged page requires a named human reviewer before it goes live. Where a figure could not be verified, the page says so instead of accepting a plausible-looking answer.

Corrections

We would rather hear about an error than not. Corrections that change a fact are made on the page, the review date is updated, and material corrections are noted on the page itself.

If you believe something here is wrong — particularly a cap, a deadline or a licensing requirement — please tell us. Those are the errors that can actually hurt someone.

What we deliberately do not publish

• Step-by-step cleanup instructions, or HowTo structured data anywhere on this site.
• Meth lab decontamination thresholds or the state lists that accompany them, which we could not verify.
• State-by-state stigmatised property disclosure law, which needs a dedicated legal research pass we have not done.
• Several widely repeated national statistics about victim compensation funding, which we could not trace to a primary DOJ or OVC document.
• Any figure for a company we have not been able to source.