What ResearchVerify is
An evidence platform built around three structural commitments.
Every COA in the database has been scored against the same rubric. No vendor pays for placement. No score is hidden. No part of the methodology is private. The platform is built around three commitments.
1. Methodology over opinion
Scoring decisions follow a published rubric, not editorial judgment. If a record scores 7.2, that number traces back through documented math to four weighted components. The rubric is at researchverify.com/methodology — anyone can read it, dispute it, or check that the math on a specific record adds up.
2. Evidence over endorsement
A high score is a high score on the rubric. It is not a buyer recommendation, a quality guarantee, or an endorsement of the vendor. The platform aggregates evidence and reports scores; what to do with that information is up to the reader.
3. Structured data over conversation
The platform itself is structured — records, scores, anomaly codes, attached notes. There is no chat, no DM system, no discussion forum on this website. Community conversation, when it happens, takes place on separately-operated channels with their own structure and rules.
How to use it
The platform serves two roles. Most people are one or the other; some are both.
Reading evidence
You want to know whether a specific compound, batch, or vendor is what the label says.
- Search by compound name, vendor, batch ID, or task number from the homepage.
- Open a record to see its score, the four scoring components, and any anomaly flags.
- Read attached community notes for context the COA alone doesn't capture.
- If a record is flagged or scores low, the methodology page documents why every flag exists and what it means.
Contributing evidence
You sell research peptides or AAS and want your COAs in the public database.
- Upload your COA at researchverify.com — drag and drop, scoring runs in about 20 seconds.
- The record appears in the public database with your batch ID, vendor name, and computed score.
- After upload, visit your record to add notes, batch context, or photo documentation.
- Your aggregate stats build on your vendor profile over time — average purity, COA count, score distribution, trend.
What ResearchVerify is — and what it isn't
The distinction matters legally and operationally. Both columns are here on purpose.
It IS
- An evidence aggregation platform for COA records
- A scoring system built on a publicly documented rubric
- A structured-annotation surface (community notes attached to records)
- Independent of any specific vendor, lab, or commercial channel
- Free to submit to; free to read
- Open to community dispute on any record or score
It is NOT
- A marketplace — no products are sold, listed, or facilitated for sale
- A sourcing channel — no contact info, links, or trade discussion
- A chat or forum — no discussion threads on the platform itself
- A regulatory body — no clinical or medical certification implied
- An endorsement system — scores are data, not recommendations
- A pay-to-rank service — no vendor payment affects scoring
Why this distinction matters
The peptide research category has had platforms that mixed structured data with chat, sourcing, marketplaces, and paid promotion. Those structures create regulatory exposure and editorial corruption. ResearchVerify keeps the platform structurally narrow on purpose: evidence in, scores out, notes attached to evidence. Discussion lives elsewhere.
The platform structure
ResearchVerify operates on two surfaces, intentionally separate.
If you want both data and conversation, you use the website plus whichever community channel you choose. If you only want data, you only use the website. Each surface operates under its own structure and terms.
Information posted on external channels does not appear in the database, and information in the database does not get cross-posted to external channels unless someone explicitly references it. The two surfaces are independent.
What's different about this approach
Three structural choices that shape every other decision.
The methodology is public, in writing
Every scoring decision — every weight, every anomaly code, every credibility modifier — is documented at /methodology. If you disagree with a scoring choice, you can identify the exact rule and argue with it. Methodology that lives in the heads of moderators is methodology that doesn't actually exist.
Overfill is treated as a buyer benefit, not a defect
Lyophilized peptide vials almost always contain more total mass than the label claims (TFA counterions, residual moisture, intentional overfill to hit active content). Buyers receive more than they paid for. The rubric scores overfill as full credit and only penalizes underfill. This is documented in detail on the methodology page; it's the single most-discussed scoring decision.
Community notes are structured, not chat
Notes attach to specific records. They aren't comment threads, replies, or messages between users. The community guide documents what good notes look like and what gets removed. The structural difference between "annotation on data" and "chat" is what keeps the platform credible.
What ResearchVerify doesn't yet do
Honesty about the current state, not aspiration about the future.
The methodology page covers limitations in detail. The short version:
- No independent sample purchasing. All COAs are user-submitted. We don't currently buy samples ourselves and send them for blind testing. Coverage is uneven as a result — popular vendors have many submissions, smaller vendors may have none.
- Limited cross-lab verification. Most records are verified against one testing lab. Cross-validation earns a credibility bonus when present but isn't the default.
- Authenticity scoring is statistical, not forensic. The Conformance Index and BS detection flag suspicious patterns; they don't issue judgments. A record can pass every check and still be fabricated.
- Compound class is heuristic. The A/B/C tier system is a simplification. Edge cases are documented per-compound.
The full list with detail is in the "What this rubric does NOT yet do" section of the methodology page.
The reference documents
The three pages that define how the platform operates.