Skip to content

Verified Transaction History

Recordkeeping register — instantiates Reputational Signal Governance

Presents a subject's completed transactions, fulfilled commitments, and defect or dispute outcomes as verified facts of record — evidence, not opinion.

Verified Transaction History is the objective-evidence layer of the pipeline. Where a rating captures what an interaction felt like, this register records what verifiably happened: deals completed, commitments met, delivery times, defect and return rates, disputes and how they resolved — each drawn from the system of record and attributed to the subject with its context. Its defining move is to trade breadth of sentiment for evidential hardness: a smaller set of facts, but facts a subject cannot self-author, so they anchor the top of the evidence hierarchy. It complements reviews rather than replacing them — the record says what occurred, reviews say what it was like.

Example

A procurement platform lists industrial components suppliers. A buyer weighing a new supplier opens its history and sees not testimonials but a ledger: ≈1,240 completed purchase orders, ≈98% delivered on time, a 1.4% defect-and-return rate, three formal disputes (two resolved in the buyer's favor). Every line ties back to actual platform-mediated orders, so the supplier can't quietly delete the bad quarter or invent the good one. The buyer shortlists on that verified delivery record and treats the supplier's own glossy self-description as the weaker signal — exactly the ranking the register is built to produce.

How it works

What distinguishes it is provenance and attribution, not solicitation:

  • Draw from the system of record. Entries come from events the platform already witnesses — payments cleared, shipments scanned, tickets closed — not from anyone volunteering input.
  • Attribute with context. Each event is bound to the subject with the details that make it interpretable: amount, date, category, counterparty type, and outcome.
  • Stamp as verified. Because these facts can't be self-reported, they form the high-quality evidence tier that outranks unverifiable claims when both appear.

Tuning parameters

  • Verification depth — platform-mediated events only, or accept attested external records too. Deeper coverage widens the picture but weakens the guarantee behind each line.
  • Attribution granularity — how much context each record carries. Richer context resists misleading comparison (a high dispute count that is really just high volume) but adds noise and can leak commercial detail.
  • Outcome inclusion — whether cancellations, refunds, and lost disputes appear alongside the wins. Showing negatives is honest and contentious in equal measure.
  • Retention window — how far back the ledger runs. Longer is fuller but risks binding a reformed subject to a distant past (leave time-decay to the decay rule).
  • Evidence tiering strength — how much more a verified fact weighs than an unverified claim when a downstream score combines them.

When it helps, and when it misleads

Its strength is that it is the hardest layer to game and the right anchor for high-stakes, low-familiarity decisions: when you don't know a counterparty and can't afford to guess, a record of what they actually did beats a wall of stars.

It misleads through what it cannot see and what it strips out. Coverage stops at the platform's edge, so off-platform conduct is invisible and absence reads as innocence; context-poor metrics invite unfair comparison; and a tidy table of percentages can lend false certainty to a record that is thinner than it looks. Its record is also easy to invoke for non-repudiation[1] — a fact the subject can't credibly deny — which makes selective presentation the classic misuse: surfacing the on-time rate while burying dispute outcomes, curating "facts" to pre-justify a conclusion. The discipline is a fixed record schema the subject can't choose per case, context shown beside every metric, and time-decay delegated to its own rule rather than hidden by quietly dropping old entries.

How it implements the components

  • attribution_and_context_record — binds each completed transaction, defect, or dispute to the subject with the context needed to read it (what, when, size, counterparty, outcome).
  • evidence_quality_weighting_rule — constitutes the verified/objective tier: facts drawn from the system of record are the evidence that outranks self-reported or unverifiable claims.

It does NOT solicit or display subjective reviews (that's Rating and Review System), aggregate its records into a score (Reputation Score or Standing Index), apply time-decay (Decay-Weighted Score Update), or detect fabricated or colluding records (Sybil, Collusion, and Brigading Detection).

References

[1] Non-repudiation — a record whose origin and integrity are strong enough that the subject cannot plausibly deny it. It is the property that lets a transaction ledger function as evidence rather than assertion, and the reason verified facts sit above self-reported ones in the evidence hierarchy. ↩