How Peptigrity actually works.

How trust scores are calculated

The Trust Score is a single number from 0 to 100 calculated from two data sources: community reviews and independent lab tests. The score scales by confidence in the data behind it — small samples carry less weight, large samples carry more.

The formula

Two weighted terms (0 to 1 each), reviews stated first:

• 0.5 × smoothed review average (normalized to 0–1 from a 0–5 scale)
• 0.5 × banded purity average (the smoothed purity % mapped to a band score, see below)

Community reviews and lab purity are each weighted at 50%: objective HPLC purity data and subjective buyer experience (quality, delivery, service, pricing) are treated as equally important signals. Neither channel outweighs the other.

Confidence (0 to 1) is calculated separately for each channel:

• conf_reviews = min(1.0, ln(reviews + 1) / ln(25))
• conf_purity = min(1.0, ln(tests + 1) / ln(15))

Each term is scaled by its own channel's confidence and the two are added. The sum is not divided back by the confidence, so low confidence in one channel costs points rather than moving that channel's weight onto the other:

• Trust Score = (0.5 × conf_reviews × reviews + 0.5 × conf_purity × purity) × 100

Because confidence is per-channel, a shop must carry meaningful volume in both reviews and lab tests to reach the top of the rankings — a large pile of lab tests can no longer compensate for a handful of reviews, and vice versa. The review channel saturates around 24 reviews; the lab channel around 14 tests.

Purity banding

The smoothed purity average enters the score as a band, not a raw percentage, so sub-1% HPLC measurement noise cannot decide rank order between shops. The precise tested percentages are still displayed everywhere on the site — only the scoring uses the band.

• ≥ 99.0% → 1.00
• 98.0–98.99% → 0.92
• 95.0–97.99% → 0.80
• 90.0–94.99% → 0.55
• < 90.0% → 0.25

Why we use Bayesian smoothing

Both the review average and the purity average are Bayesian-smoothed using priors. A new shop with one 5-star review does not get treated the same as an established shop with 150 reviews averaging 4.8. The smoothing uses:

• Review prior: 4.2 stars (the platform average)
• Purity prior: 98.5% (the platform average)
• Smoothing constant: K = 8

The smaller the sample, the more it's pulled toward the prior. Once a shop accumulates real volume, its actual data dominates. This prevents small-sample manipulation: a single cherry-picked review or lab test can't carry a brand-new shop to the top of the rankings.

Why K = 8? A higher smoothing constant dampens small-sample volatility as the platform's overall dataset grows. With more reviews and lab tests on file, the cost of letting a single 5-star or 1-star outlier shift a shop's score has gone up — K = 8 makes new shops earn their rank from a larger body of evidence before pulling away from the platform average. The trade-off is that very-high-purity shops with only a handful of tests rank slightly lower than they would under K = 5; that's the intended behaviour.

Verification tiers

Every shop carries one of these labels based on data depth:

When one data channel has no trustworthy signal, its term contributes 0 — so the score caps at the other channel's weight (50 for reviews-only, 50 for lab-only — both single-channel tiers are treated identically). Once both channels contribute, the score can reach 100.

What the Trust Score does not include

The Trust Score does not factor in shop age, listing date, product selection, pricing, website design, social media presence, advertising spend, or any payment to Peptigrity. The calculation uses only two data inputs: review ratings and lab test purity data.

Some reviews and lab tests are left out of those two inputs:

• Reviews hidden through the review-dispute process. Reviews marked unverified stay visible and still count.
• Lab tests that have not been approved yet.
• Lab tests quarantined because the laboratory that ran them could not be confirmed. They are kept, not deleted, and can count again once the laboratory is established.
• Lab tests with no purity figure, such as net-content-only or sterility certificates. They are not part of the purity average or of the test count that sets the lab channel's confidence.
• 1 or 2 reviews on a shop that has lab tests. Reviews start counting at 3 (see Verification tiers).

And these are not inputs at all: whether a shop is claimed, shop board threads, group buys, moderation actions, reputation points and researcher tiers, badges, a certificate's confirmation status, and who sourced the tested vial. Reviews where the shop offered an incentive count like any other review (see Reviews the shop offered something for).

Score updates

Trust Scores recalculate automatically when new reviews or lab tests are published. A single new data point can change a shop's Trust Score — though Bayesian smoothing and confidence scaling mean the change is proportional to the existing data depth. A new review for a shop with 200 reviews moves the score less than a new review for a shop with 3 reviews.

Sub-ratings (independent of Trust Score)

Beyond the Trust Score itself, every Peptigrity review breaks down the buyer's experience into four sub-rating dimensions on a 1–5 scale: Quality, Delivery, Service, and Pricing. These are averaged separately per shop and displayed on each shop profile. The sub-ratings are informational — they help buyers see exactly where a shop performs well or struggles — and they are not factored into the Trust Score calculation itself.

A separate Would Buy Again indicator (yes/no) is tracked on every review and displayed as a percentage on each shop profile. This is also informational and not factored into the Trust Score directly.

Methodology updates

The Trust Score formula evolves as the platform matures. Updates are listed here in reverse chronological order so any change to the calculation is visible alongside the rules it produced.

• June 2026 — Review/purity weighting rebalanced to 50/50; confidence split per channel. Community reviews now carry equal weight to lab purity, reflecting Peptigrity's review-first mission. Confidence is now calculated separately for the review channel and the lab channel, so a shop must have meaningful volume in both to reach the top of the rankings — no shop can rank highly on lab tests or reviews alone. Purity is now scored in bands to prevent sub-1% measurement differences from affecting rank.
• May 2026 — Bayesian smoothing constant raised from K = 5 to K = 8. Higher K dampens small-sample volatility further; as the dataset has grown, the case for trusting a few-data-point shop's raw average has weakened. Same formula, stronger pull toward the prior for shops with fewer than ~10 reviews or lab tests.

Review policies

Who can submit reviews

Any registered Peptigrity user can submit 1 review per shop. Registration requires a username, email address, and password. Email addresses are never displayed publicly.

What reviews include

Each review includes 5 sub-ratings on a 1–5 scale:

1. Overall — general satisfaction with the purchase experience
2. Quality — product purity, consistency, and packaging
3. Delivery — shipping speed, tracking accuracy, and packaging integrity
4. Service — customer support responsiveness and helpfulness
5. Pricing — value relative to product quality and market rates

Each review also includes a "Would buy again" indicator (yes or no) and a free-text description for each sub-rating category.

Reviews the shop offered something for

Shops may offer a discount or reward for leaving a review on Peptigrity. The offer must apply to any honest review, not to a positive one, and the shop must tell the customer to tick the "offered a discount" box on the review form. Incentivized reviews are tagged. They currently count toward Trust Score like any other review; Peptigrity will revisit this as review volume grows. Undisclosed incentives are a guideline violation and reviews may be removed.

The disclosure is the reviewer's, not the shop's: it is a checkbox on the review form and nothing else sets it. Peptigrity does not infer it from review text, timing, or a shop's promotion history, so a review without the tag means nobody said there was an incentive — not that one was ruled out.

Review moderation

Reviews are moderated for 4 criteria before publication:

1. Relevance — the review relates to the shop being reviewed
2. Registration — the reviewer has a registered Peptigrity account
3. Prohibited content — the review contains no personal information, threats, hate speech, or illegal content
4. Duplication — the reviewer has not already reviewed this shop

Reviews meeting all 4 criteria are published without editing. Negative reviews are not suppressed, deprioritized, or modified.

Lab test policies

What lab tests include

Each lab test submission includes the peptide name, brand or shop name, HPLC purity percentage, stated quantity (label), tested quantity (actual), endotoxin level (when tested), testing laboratory name, test date, batch ID (when available), and an image of the Certificate of Analysis.

Submission and verification

Lab tests are submitted by registered community members through the submission form. Each submission is reviewed by the Peptigrity team before publication. Peptigrity does not modify purity values or quantity measurements.

Publication transparency

All lab test results are published transparently regardless of outcome. A lab test showing 85% purity receives the same treatment as a lab test showing 99% purity. Low purity results are not hidden, delayed, or deprioritized.

Testing laboratory independence

Peptigrity publishes lab tests performed by third-party testing laboratories with no financial relationship to the peptide shop whose product was tested. In-house testing by vendors is not accepted.

What current lab data can and cannot tell you

Most lab tests currently in the database are independent third-party Certificates of Analysis that Peptigrity collects from shops' official channels and reviews before publishing. Because shops choose which COAs to publish, these tests reflect the batches a shop has made public and may not represent every production batch. Lab test pages show the certificate's confirmation status (see below) and, where it is recorded, who sourced the tested vial (shop-provided, bought at retail by a member, or funded through the Community Testing Grant), so a reader can weigh each result for what it is.

Certificate confirmation

Every certificate carries a confirmation status that says whether it has been checked against the testing laboratory's own record. The same wording is used on lab test pages and in the vial lookup:

• ✓ Lab-verified — the laboratory runs a public portal that looks up individual reports, and the certificate carries the report number or verification code to find it there.
• Unverified — not yet checked against the laboratory's portal. This describes Peptigrity's queue, not the shop or the certificate.
• Lab has no portal — the laboratory publishes no public lookup, so the report cannot be confirmed independently. This is not a mark against the certificate or the shop.
• Report no. not found — the laboratory's portal returned no record for the report number on the certificate.

A certificate whose figures do not match the laboratory's own record is removed from the vial lookup and from fact-sheet counts. Confirmation status is information for the reader: it does not change a shop's Trust Score.

Lot numbers

The lot or batch number printed on a vial can be looked up at peptigrity.com/coa. The lookup ignores capital letters, spaces and punctuation, so "06-2026-BPC-10" and "062026bpc10" find the same certificates. Results are grouped by shop and never merged, because resellers can share a lot number and one shop's measured figures must not be attributed to another shop's vial.

Some values typed into the batch field cannot identify a batch, and the lookup refuses them rather than guess: placeholders such as "N/A", values with no digit in them (cap colours such as "Black Cap" are shared by many shops), and codes shorter than 5 characters. Nothing is deleted: the value as submitted still shows on the lab test page. A certificate describes the vial from that lot which the laboratory tested, not every vial in the lot.

Imported certificates

Some certificates are imported from a testing laboratory's public certificate portal, starting with Freedom Diagnostics. These were published on the portal by the laboratory's client, usually the shop, and were not submitted to Peptigrity. Each one is read and checked by Peptigrity staff before publication and then goes through the same approval step as any submitted lab test. A certificate is skipped when Peptigrity already holds the same report number, verification code, or the same lot from the same shop. Imported certificates are recorded as shop-provided, and each one's certificate record says it was imported from the laboratory's public portal.

A portal holds only the certificates its clients chose to publish. Results that a shop preferred to keep private are not there, so imported certificates lean toward results the shop was willing to make public.

Measured quantity against the label

Where a certificate reports both the labelled quantity (what the vial says it contains, for example 10 mg) and the measured quantity (what the laboratory found), Peptigrity shows the measured figure as a percentage of the label: measured ÷ labelled × 100. It describes the one vial the laboratory tested.

Gross and net content. Laboratories do not all weigh the same thing. Some report the gross mass of the powder in the vial, which includes salts and moisture as well as peptide. Others report net peptide content only. A label normally states the peptide amount, so a certificate that weighs gross mass against a net label can read well over 100% for an ordinary vial. Certificates often do not say which basis they used, and where it is not known the figure is shown as printed with the basis marked as not stated.

Why a figure above 110% is not shown as a good result. Peptigrity cannot tell a gross-mass reading from a real overfill, and an overfill is a filling inconsistency rather than a merit. Readings above 110% are common: in September 2026 about 40% of single-compound certificates that report both figures were above it. So a figure between 95% and 110% is shown as in line with the label, a figure below 95% is marked as under the label, and a figure above 110% is never coloured as a good result: the lot badge prints it in plain text and the lab test listing marks it for attention.

Blends are excluded from every label-against-measured figure that summarises more than one certificate, such as a shop's median. A blend certificate's quantities can describe the whole vial or a single component, so its percentage is not comparable with a single-compound vial's. Certificates whose figures did not match the laboratory's own record are also left out.

What “passed” means on a certificate

A lab test passes when every result its certificate reports is inside its limit. The Trust card badge counts passed tests with this rule, and the per-batch certificate badge uses the same rule to decide whether it may render in a result colour. It is a count of results against a published definition, not a grade of the shop.

• Purity. The certificate reports an HPLC purity figure. No minimum purity is applied yet; when one is set it will be stated here, and it will apply to every test at once.
• Endotoxin, where the certificate reports it: the result is within its limit. A result the certificate prints as over its limit, failed or out of specification fails the test.
• Heavy metals, where the certificate reports them: negative or within limits.
• Sterility, where reported: a pass. Lab test submissions do not yet record sterility as its own result, so no test currently passes or fails on sterility.

A certificate that reports only purity passes on purity alone: a result that was not reported is not counted as a failure, and not as a pass either. A result written in a way that cannot be read as a pass or a fail is not counted as a failure. A certificate whose figures did not match the laboratory's own record is left out of both the number of tests and the number that passed. Shops cannot change how a test is counted.

Shop listing policies

How shops are listed

Shops are added to Peptigrity through 2 methods: community submission (a registered user adds the shop) or editorial addition (the Peptigrity team identifies the shop). Listing is free. No shop pays a listing fee, commission, or subscription.

Shops without a website

A shop does not need a website to be listed. A shop submitted without one needs a name and at least one contact channel, such as Telegram, WhatsApp, WeChat or Signal, and appears once the Peptigrity team has reviewed the submission. Its page lists those channels in place of a website link. A channel is marked Stated when the shop or a submitter typed it in and nobody has checked it, and Confirmed once the shop has proved control of it during a claim. WhatsApp and Signal numbers are shown with only the last four digits. A shop without a website is scored with the same formula as every other shop: no penalty and no bonus.

Shop verification status

Each shop carries 1 of 2 verification statuses:

1. Community Submitted — the shop was added by a community member or the Peptigrity team. The shop owner has not verified their identity.
2. Claimed — the shop owner has contacted Peptigrity, proved ownership, and actively manages their listing. Claimed shops can update their profile and respond to reviews.

Claiming a listing does not affect the trust score. The verification status indicates only whether the shop owner is actively managing their profile.

Claiming by channel

A shop with a website is claimed by showing control of the site: an email from an address on the shop's domain, or a page on the site that links to the claimant's Peptigrity profile. A shop without a website, or one that prefers it, can be claimed by proving control of a contact channel instead. The claimant picks one of the shop's listed channels or adds the one they control, posts the fixed text "Peptigrity verification" in it from an admin account, and submits the claim. The text is the same for every shop: there is no code and no bot. A member of the Peptigrity team opens the channel, looks for the post and confirms by hand. WeChat groups cannot be opened from outside, so for WeChat a link to a screenshot is accepted.

A confirmed claim marks that channel Confirmed and gives the account the Shop owner tag. Up to 5 accounts can manage one shop; the first owner can invite others with a one-time link that expires after 7 days. Claiming by channel carries the same rule as any other claim: it changes nothing about the Trust Score, the ranking or which reviews are shown.

Ranking order

Shops on the directory page are ranked by trust score in descending order. The shop with the highest trust score appears first. No shop can pay for priority placement, featured positioning, or ranking modification.

Fact sheets

Each shop, testing laboratory and compound page opens with a fact-sheet paragraph. It is generated automatically, not written by hand: every figure in it comes from the Peptigrity database (lab tests on file, the laboratories that ran them, reported purity and label figures, community reviews, contact channels and whether the shop is claimed), and it ends with the date the figures were read. A missing value is stated rather than left out: "not stated" means the shop or laboratory has not provided it, and "none on file" means Peptigrity's own records hold nothing. The same paragraph is used for the page's search description and its structured data, so the three cannot disagree.

Removal policy

Shops cannot request removal from Peptigrity. Once listed, a shop remains in the directory along with all associated reviews and lab test data. The only exception is a listing created in error.

Promotion policies

Promotions are available to shops with a Trust Score of 70 or higher. Each promotion is reviewed by the Peptigrity team before publication. Promotions display the shop's current Trust Score and, when available, the most recent lab test purity for the promoted peptide.

Expired promotions remain visible on the platform for transparency and historical reference.

A claimed shop's owners can also post promo threads on the shop's own board. Those threads are labelled Promo, are ordinary forum posts under the community rules below, and are the only place in the forum where a shop may post promotions, store links or discount codes. Board promos are not a ranking input.

Content moderation

Peptigrity moderates 3 types of content: reviews, lab test submissions, and discussion posts. Moderation applies the same standards to all content regardless of the shop or user involved.

Community rules

Discussions, shop boards and group buys are for adults (18+). Members may discuss any shop, compare vendors and organise group buys. The rules:

1. Compound names only. Brand names are not used for products on this platform.
2. No controlled substances. Discussion of hormone therapies is allowed; sourcing them is not.
3. Shop owners are welcome in every thread and are always tagged. Promotions, store links and discount codes are permitted only in the owner's own board and on promotions; elsewhere they are removed.
4. Group buys are coordination only. Peptigrity handles no money, holds no funds and vouches for no organizer.
5. No personal information, threats, hate speech or content involving minors.
6. Reviews belong in the Reviews tab of the shop page, not in threads.

Content is removed only if it breaks one of those rules. Reviews and Scam-watch threads are checked before they publish; everything else in the forum goes live immediately and is moderated afterwards.

How moderation works

In the forum and on shop boards, Peptigrity staff may hide a thread or reply, lock a thread, move it to another category, warn a member and ban an account. Members who have reached the Principal investigator tier may take the same actions except banning, which is staff only. A shop owner may hide replies on their own shop board and nowhere else. Hidden content is taken out of view, not erased, so a mistaken action can be reversed. Every one of these actions is published on the moderation log with what was done, to which thread, reply or member, the rule it was taken under, who took it and when. Email addresses are never shown there. Moderation changes what is visible; it never changes a Trust Score, a ranking or the order of any list. Reviews are moderated separately, under the review policies above.

Reputation and ranks

Reputation points measure participation, not authority. They are earned by contributing — reviews, lab tests, protocol logs, discussion posts — and they move a member up a ladder of ranks shown on their profile and beside their name in threads.

Reputation has no effect on Trust Scores, rankings, or how any shop is presented. A high-reputation member's review counts exactly as much as a new member's. Points are not currency, cannot be bought, transferred or gifted, and unlock no editorial privileges.

Shop owners earn points and tiers under exactly the same rules as every other member, including for threads on their own shop board. Claiming a shop awards no points and no tier, and the Shop owner tag is shown beside an owner's tier, never in place of it.

Ranks

A rank is derived from the point total at the moment you look at it. No rank is stored anywhere and none is awarded by hand, so a rank can never be stale, granted as a favour, or out of step with the points behind it.

The progress bar on a profile measures progress within the current rank, not toward the next rank's threshold from zero. A member on 47 points is 37 of the 40 points through Contributor — 93% — rather than 47/50 of the way to Regular.

What earns points

Three of these are deliberately hard to farm. Lab test points are paid on approval, never on upload, and a failing result earns exactly what a passing one earns — paying more for bad findings would pay people to invent scandals. A log entry pays once per the log's own interval, so a daily logger cannot out-earn a weekly one by splitting a week into seven rows. And a completed log only counts as well-documented — and earns its points — at 5 entries or more.

Points are awarded when a contribution is made and are not currently deducted if that contribution is later removed. Reputation is therefore a record of participation over time rather than a live count of what is still standing — another reason it is kept away from anything that ranks a shop.

Researcher tiers

Researcher tiers reflect how much someone has contributed across reviews, discussions, and submissions, and how long they've been around. They are not a verification of purchases or identity. A tier never affects a shop's Trust Score.

A tier is not the same thing as reputation points above it. Points are a running total that only goes up. A tier is a description of an account as it currently stands: it is computed from criteria rather than a balance, every condition above the first tier has to be met at once, and it goes back down if the reviews or replies behind it are removed. That is deliberate — it means no single repeated action carries anybody up the ladder, and it means the tier beside a name is true today rather than true once.

Tiers are recomputed nightly, and immediately for you when you publish a review, a reply or a submission — so a first review shows its tier straight away rather than the next morning. Peptigrity staff accounts hold no tier; they carry a team mark instead.

Questions

For questions about these rules or to report a policy concern, visit the Contact Us page or review the FAQ.