START HERE / NO SCIENCE DEGREE REQUIRED
Fewer tabs.
Better questions.
A beginner’s guide to compounds, research and COAs. Short answers first. Sources when you want to go deeper.
Public educational guide · Updated 8 October 2026 · No account needed.
1. Put the picture together
A claim, a study and a COA answer different questions.
A claim is what someone says a compound can do. A study tests a specific question. A COA records laboratory findings about a particular sample. A community discussion records what people say they experienced.
You need to compare these pieces, rather than letting one stand in for all the others. A high purity result cannot prove a health benefit. A promising study cannot authenticate a supplied vial.
Three questions to take away
- What exactly is the claim?
- What evidence directly supports that claim?
- What remains unknown?
2. Learn the basics
Start with the name, form and purpose—not a protocol.
A compound is a chemical substance. A peptide is a chain of amino acids; peptides are one group within the wider compound catalogue. A blend contains more than one component. Similar names do not make products interchangeable.
Identify the exact substance, formulation and route studied. A skin cream and an injected product are different situations. “Research use only” is a label, not evidence that a product is approved or suitable for use in people.
Three questions to take away
- Is this one compound or a blend?
- Does the formulation match the research?
- Am I looking at a draft profile or an established finding?
3. Turn a headline into a question
“Anti-ageing” is too broad to evaluate on its own.
Rewrite the headline as a specific question: in which people, using which formulation, compared with what, over what period, and measured how?
For example, “improves skin” might mean a questionnaire score, an instrument measurement or a photograph. Those outcomes are different. Ask for the original study rather than accepting a screenshot of a conclusion.
Three questions to take away
- Who was studied?
- What changed, and how was it measured?
- Was the effect meaningful or just described as significant?
4. Read a study without reading every word
Look at the methods and results before trusting the conclusion.
Start with the study question, participants, comparison group, formulation, duration and outcomes. Then look at the size of the effect, uncertainty, unwanted effects and limitations.
Cell and animal studies can suggest mechanisms. They do not establish the same benefit in people. Human trials address more direct questions, but a small or poorly designed trial may still leave substantial uncertainty. Reviews depend on the quality and relevance of the studies they include.
Check whether the result came from random allocation and blinding, whether people dropped out, and whether the authors declared funding or conflicts. A registered trial is not a completed, positive trial.
Three questions to take away
- Cells, animals or people?
- Was there a useful comparison group?
- Does this outcome support this exact claim?
5. Match the batch to the report
A genuine report can still be the wrong report for a product.
Find the lab name, report identifier, sample description, batch or lot, dates and complete pages. Compare identifiers with the product information. A missing identifier means the link has not been established.
Use the issuing laboratory’s official verification service when available. A supplier screenshot or QR code alone does not establish authenticity: check the destination and compare the lab’s record. Even a verified report does not prove that every supplied item matches the tested sample.
Three questions to take away
- Does the batch match?
- Can the issuer confirm the report?
- Who submitted the sample, and is its handling documented?
6. Understand HPLC—and the other numbers
Purity, identity and quantity are separate results.
HPLC separates components for analysis. A reported purity percentage depends on the method and detector. It is not automatically the percentage of the vial’s total weight that is the desired peptide. Water and counter-ions can contribute to gross weight; peptide content is a separate concept.
Identity testing asks whether the measured material is consistent with the intended compound. Quantity testing asks how much is present. Concentration includes a volume unit, such as mg/mL. Do not compare it directly with a total amount in mg.
Sterility, endotoxins and contaminants need their own tests. “99% purity” does not establish these. Look for the method, result, units, specification and limitations for each reported test.
Three questions to take away
- Which tests were actually performed?
- Is this purity, quantity or concentration?
- What was not tested?
7. Read your provisional score
The number measures documented evidence, not product safety.
Our provisional score summarises the evidence our system can identify in a COA. It is separate from the purity percentage printed by a laboratory. Unreadable or unconfirmed information earns no points; that does not mean a laboratory test failed.
Automatic assessments are capped at 70/100 while issuer verification is unchecked. Read the short summary and critical flags first. Members can open the reasons and keep private reports. Compare extracted numbers with the original PDF.
A higher documentation score cannot confirm suitability for use, clinical effectiveness or that your supplied item is the tested sample.
Three questions to take away
- What supports the number?
- What needs confirmation?
- Could an extraction error explain a missing result?
8. Use experiences as context
Useful observations are different from proof of an effect.
People’s accounts can help identify questions worth asking. They cannot reliably separate an effect from other changes, expectations or coincidence. Popularity and repeated testimonials do not replace controlled research.
If you record observations, keep the goal, dates, formulation, actual exposure and other changes clear. Record no change and unwanted effects too. Compound Society offers private tracking; it does not prescribe a protocol or publish your records automatically.
Chain Collective is part of this project and hosted separately on Telegram. Community and supplier links are disclosed on our community page. We receive no commissions, affiliate payments or referral fees from these external links.
Three questions to take away
- What else changed at the same time?
- Is the account specific enough to interpret?
- Does anyone have a commercial interest?
9. Know the next useful step
An unanswered question is a useful finding.
Missing evidence is a reason to pause and clarify, not fill the gap with a confident guess. Ask the lab about a report, read the original study, or speak with a qualified clinician about personal health decisions.
You do not need to be using a compound to join or learn. Begin with a research question, learn the vocabulary and compare what is claimed with what is documented.
Three questions to take away
- Which source can answer my remaining question?
- What would change my view?
- Do I need professional advice rather than another discussion thread?
CLAIM → EVIDENCE → LIMIT
“GHK-Cu improves skin.” What does that mean?
This is a worked reading exercise using one study, not a complete review of GHK-Cu.
- The question
- Did topical GHK-Cu skin care improve outcomes after CO₂ laser resurfacing?
- The study
- A randomised study published in 2006. Thirteen patients completed it. Outcomes included redness, wrinkles, skin appearance and questionnaires.
- The finding
- The authors found no significant difference between groups on the objective redness, wrinkle or skin-quality assessments. Patient-reported satisfaction was higher with GHK-Cu.
- The limit
- This small study in a specific post-treatment setting does not establish broad anti-ageing, hair or recovery claims, or effects from another route or formulation.
- The COA question
- A report could describe the tested sample’s chemistry. It cannot establish those clinical benefits.
Your comparison checklist
Use this as a reading aid. Ticking everything does not make a product safe. Checks stay here only while this page is open.
0 of 8 reading checks completed
Put it together here.
Explore draft profiles and their sources, check a PDF, and find the community. Members can read detailed COA reasons and keep private observations. Research profiles remain drafts; these tools do not independently verify suppliers.