How We Source Benefits
Every live page has to survive a simple test: can a careful reader see why we say a benefit is real—and who should skip it?
Our rules
- Primary over pile-ons — Prefer NIH, CDC, NIST, journals, or the vendor’s own dated docs over roundup blogs.
- Access dates — Every claim we lean on gets a source link and the date we checked it.
- Who it’s not for — Mandatory. If we can’t name exclusions, the page isn’t ready.
- Affiliates don’t write the verdict — Labeled links never buy a higher benefit or hide a caveat.
- Ask → queue → gate → publish — Typing a question does not instantly create an indexed URL. Drafts need sources; health topics wait for review.
- Corrections — If a source changes or we got it wrong, we update dateModified and the Sources list.
What we never do
- Bought rankings or “#1” badges
- Cure or disease-claim language
- Undisclosed paid placements
- Auto-indexing thin AI clones
FAQ
Can I ask for any topic?
Yes via /ask/. Only the wedge allowlist can publish. Everything else stays queued.
Do you use AI to write?
For outlines and first drafts only. Named sources and human edit gates decide what goes live.
Follow the spider web
Live topics cluster on the benefits hub. Start from hubs like AI tools, heat pumps, standing desks, or reading — each page links 3–5 related receipts.