Skip to content

Customer Trust Principles

Trust is treated as a design principle, not a marketing slogan. These principles apply to every venture and every product, regardless of how small.

Accuracy

Any factual, calculated, or informational output a product presents to a customer should be correct to the best of the company's ability to verify it. Where a product performs calculations (pricing, nutrition, inventory shelf-life estimates, planning outputs), the underlying logic should be documented and testable, not just "it looked right in a quick check."

Transparency about limitations and uncertainty

Products should be honest about what they don't know or can't guarantee. If a feature is an estimate rather than a measurement (e.g., an estimated shelf life rather than a lab-verified one), say so in the product itself, not just in fine print. Do not let a UI imply more certainty than the underlying data supports.

Distinguishing verified fact from estimate

This mirrors the evidence categories in docs/company/research-and-evidence-principles.md, applied to customer-facing content: a verified fact (e.g., "this product contains X grams of sugar" from a source label) must be presented differently from an estimate or inference (e.g., "this item is likely past its best-by window based on typical shelf life"). Blurring the two erodes trust the first time a customer notices the difference.

Human review for high-risk claims

Any customer-facing claim in the following domains requires human review before shipping, regardless of venture:

  • Legal claims (terms of service, compliance statements, warranties)
  • Financial claims (pricing guarantees, savings estimates, tax-related statements)
  • Privacy claims (what data is collected, how it's used, how it's protected)
  • Security claims (see SECURITY.md — in particular, do not conflate workbook/sheet protection with actual security)
  • Medical or health-adjacent claims
  • Nutrition claims
  • Food-safety claims (particularly relevant to Shelfery — expiration estimates, storage guidance, spoilage risk)

AI-generated drafts of this content are allowed; AI-approved final versions of this content are not.

No dark patterns

Products must not use manipulative design to extract behavior a customer wouldn't choose with full information: no fake urgency, no confirm-shaming, no hidden recurring charges, no artificially difficult cancellation flows, no pre-checked opt-ins for things that benefit the company more than the customer. This applies across all ventures regardless of platform (app, web, marketplace listing).

Honest marketing claims

Marketing and product-listing copy may only describe capabilities that are actually implemented and shipped, not roadmap items or aspirational features. See docs/company/commercialization-principles.md for the fuller commercialization-specific rules this implies, including handling of social proof and platform-specific disclosure requirements.

Why this document exists separately

Trust principles are broad enough to apply across product philosophy, evidence handling, and commercialization, but distinct enough to deserve their own reference point — this document is the one to link to when a specific product decision needs to be checked against "does this respect the customer's trust," rather than re-deriving the answer each time from more general principles.


Report an issue about this page