The BackerGuardian Trust Meter scores crowdfunding campaigns 0–100 across 5 weighted dimensions: Creator Credibility (30%), Product Authenticity (25%), Technical Feasibility (20%), Supply Chain Risk (15%), Communication Quality (10%). Kill-switch conditions cap any score at 20/100 when critical fraud indicators are detected. This page renders methodology v3.0 — the exact rule set the scoring engine runs, also published as machine-readable JSON at /methodology.json.

How the Trust Meter Works

We don't hype. We measure trust. Every signal below is applied by the same validator-checked engine — this page cannot drift from the algorithm because it is generated from it.

The five dimensions

Each dimension starts at a baseline and moves with the evidence-backed signals listed below. Positive signals require evidence too — a score is earned in both directions.

Creator Credibility

weight 30% · baseline 50
-50Previous campaign failed to deliver (rugpull)
-30Previous campaign delivered >1 year late with poor communication
-25Previous failed/canceled campaign not addressed in current campaign
-20Zero-Zero profile (0 created, 0 backed)
-10No external social verification (LinkedIn/website/company registration)
+10Backer history > 5 projects (part of the community)
+10Verified real identity (established professional/company history)
+30Successfully delivered 1+ campaigns to satisfied backers

Product Authenticity

weight 25% · baseline 50
-50Resale/rebrand of an existing product (short of verified pre-launch listing, which is a kill switch)
-40Only 3D renders shown, no photos of a physical unit
-20Prototype looks materially different from renders
-20Demo video fully CGI or heavily edited with no raw footage
+10Independent third-party hands-on review exists
+10Factory/tooling photos showing production preparation
+30Raw, uncut video demo of prototype functioning as claimed

Technical Feasibility

weight 20% · baseline 70
-30Too-good-to-be-true specs vs comparable products
-20Funding goal under ~50% of estimated BOM/tooling cost
-10Aggressive timeline (<3 months from campaign end to hardware delivery)
+10Standard/off-the-shelf components lower execution risk
+20Clean explanation of specific technical challenges and solutions

Supply Chain Risk

weight 15% · baseline 60
-20No manufacturing partner named
-20Shipping rates undefined or charged later with no estimates
-15Flexible funding with no production roadmap
-10Complex customization options create SKU-count risk
+20Named logistics/fulfillment partner
+20Photos of molds/tooling creation

Communication Quality

weight 10% · baseline 70
-30Creator silent in comments (0 replies to backer questions)
-20Word-salad updates: vague philosophy, no concrete progress
-10Deleting negative comments
+10Detailed FAQ covering refunds and warranty
+10Active creator participation with technical depth

Kill switches

Some findings end the conversation. Any one of these caps the total score at 20/100 regardless of every other dimension:

  • Reselling/Arbitrage: product listed on AliExpress/Alibaba/Temu dated prior to campaign launch
  • Identity fraud: creator photo is stock or AI-generated
  • Creator linked to banned account or known scam ring
  • Claims violate basic physics

Interpreting the score

Score RangeRatingMeaning
80–100Low RiskLow risk profile - proceed with standard diligence
60–79Medium-LowModerate risk - due diligence recommended
40–59MediumElevated risk - exercise caution
21–39Medium-HighHigh risk - significant concerns
0–20High RiskCritical risk - avoid

The monetization firewall

Scores are provably independent of money, by construction: monetization eligibility is computed from the score (threshold 60/100), never the reverse. Campaigns below the threshold have all affiliate links stripped automatically, and a validator blocks any record that claims eligibility its score doesn't earn. We disclose in both directions: some of our highest-scoring campaigns carry no affiliate relationship at all, and coverage is selected by discovery criteria — never by commission availability.

Transparency

Analysis combines automated checks (reverse-image search, platform data, marketplace scans) with human judgment; every scored record carries a dated verdict and, on current-standard records, inline source links. The full rule set behind this page is public at /methodology.json — if we change the algorithm, that file changes, visibly.