Community review

The platform commits to two things that are in tension: anyone can contribute data, and the data is rigorous enough for professionals to trust. The review process is how these coexist.

The long tail of environmental data is too vast for any single database to cover, so open contribution is necessary. But LCA values don't naturally converge — they depend on system boundaries, allocation choices, and regional context that non-specialists routinely get wrong, in ways that look defensible but aren't. So contribution without review is a credibility liability. Review is the mechanism that makes open contribution add up to professional-grade data.

What a contribution is

Any user-submitted unit of data — a new activity, a new flow, or an update to an existing one. A contribution carries the data itself, source citations, and a rationale. Accuracy scores are computed, never entered: a component that cannot be measured scores 0.5 rather than a guess. Contributors are encouraged to cite sources but not required to.

A contribution moves through a small set of states:

unreviewed → under review → approved / rejected / needs changes
                                     ↓
                                 published        ← visible in default calculations
                                     ↓
                                 contested        ← active dispute
                                     ↓
                                 re-reviewed

Only approved and published contributions count as reviewed for accuracy scoring and for visibility in high-stakes surfaces.

Three tiers of scrutiny

Not every contribution needs the same level of review. The tier is assigned automatically, based on the contribution's significance.

Tier	Applies to	Reviewers	Turnaround
Lightweight	Contributions that displace nothing, aren't regulated, and have low downstream visibility. e.g. a new artisan-bakery activity in a region not yet covered.	One reviewer	A few days
Standard	Most contributions — new activities others may use, updates that change accuracy scores. e.g. a corn-farming activity where one already exists.	Two reviewers (a third adjudicates when they disagree)	A week or two
High-stakes	Anything that appears in company comparisons, sits in a regulated category (food labelling, building materials, fuel), or contests a published number.	Three reviewers incl. a credentialed LCA professional, plus a public comment period	Weeks

A reviewer who encounters a contribution that warrants more scrutiny can escalate it.

Who reviews

Reviewer status is granted, not assumed, and there are three ways in:

Credentialed — verified LCA credentials (certified-professional status, peer-reviewed publications, recognised-society membership) grant full reviewer rights immediately.
Reputation-based — a track record of accepted contributions earns an invitation to apply, granting reviewer rights in the categories where expertise is demonstrated.
Domain-expert — a chemist, transport engineer, or hydrologist can review contributions in their domain without being an LCA practitioner.

Level	Who	Can review
General	Reputation-based, broad LCA background	Lightweight & standard, in approved categories
Domain	Domain-expert entry	Lightweight & standard, in their domain
Senior	Credentialed LCA professional	All tiers
Adjudicator	Senior reviewer on dispute panels	Dispute resolution

Reviewers carry a reputation score that rises with accurate, consistent reviews and falls with ones later overturned. Status can be suspended or revoked for conflicts of interest, systematic inaccuracy, or misconduct — and revocation is recorded, not erased.

What a review does

A reviewer's job is to confirm that the activity and flow declarations are methodologically sound and that the cited sources support them. Accuracy scores are not set by hand — they are computed from the standing of the organisation whose data it is, and from geography and year — so a review changes a score only by changing what it measures. Reviewers read the contribution, inspect the declared records, and evaluate the cited sources. Each review records one of three decisions: approved, rejected, or needs changes. Reviewers must declare any conflict of interest before reviewing.

Approval can also elevate an organisation's standing to peer-reviewed or verified, which raises the source component of every input it publishes.

Disputes

Because the platform surfaces head-to-head comparisons with real commercial consequences, contested numbers get a structured path. A dispute can be raised by a reviewer, by the subject of a contribution, or by a qualified third party, and it can challenge the declarations behind a score, the methodological soundness of a declaration, or an alleged conflict of interest in the original review.

Disputes go to a panel of senior-reviewer adjudicators, who uphold, modify, or overturn the original decision. While contested, data stays usable in calculations but carries a visible contested flag, and consumer comparison surfaces hold it back from head-to-head views until resolved. Repeat disputes without new evidence face a higher bar — disputes can't be used to suppress valid data.

Keeping data honest over time

Every review action is auditable — who reviewed what, when, what they decided, what conflicts they disclosed. A published result can be reconstructed as it was understood at the time, even if reviews later change.
Data ages. The temporal component measures the gap between an input's year and the activity's year, so a contribution scores lower as the activities that use it move on — which is the prompt to re-review it.
Reviews are open by default. Reviewer name and credentials are visible; anonymity is opt-in with cause (and always auditable internally), because anonymous reviews are less defensible to outsiders.

The process is deliberately formal, because the platform's professional credibility depends on it.