Research methodology
Preview · calibration ongoing. These are the high-level principles of the 3D Commerce Index, not a claim that its methodology is finalized.
- Methodology status
- Calibration ongoing
A public aggregate methodology version has not yet been assigned.
Fixed cohort and sampling origin
The current completion audit did not redraw or optimize the cohort after seeing capability results. It reused the exact 23 companies successfully collected in the earlier Dutch furniture pilot. That pilot began from CBM’s Woonmeubel member category captured on 17 August 2026: 93 category members were recorded, 71 remained eligible after the documented exclusions, and a deterministic selection produced 24 companies. One selected company, HML Bedding, could not be collected successfully and was retained as a collection failure rather than replaced, leaving the fixed 23-company reporting denominator used here. These percentages describe this cohort and are not a market-wide prevalence estimate.
Detected
A detected capability requires reviewed qualifying evidence supporting presence within the audited snapshot. Detection is evidence of presence, not a guarantee of continuous future availability.
Not detected in our audit
An audited non-detection requires a separate absence assessment with sufficient automatic coverage, explicit human terminology and scope review, generated negative evidence, and evidence approval. It is scoped and time-bounded and must not be rewritten as proof of absolute absence.
Inconclusive
Inconclusive means the reviewed evidence does not support either a detected state or a defensible audited non-detection. Coverage gaps and taxonomy-boundary cases remain inconclusive rather than being forced into a negative result.
Human review and publication approval
Research findings require human review. A completed automated collection run is not publication approval. Positive evidence, absence-review decisions and generated negative evidence remain independently reviewable gates.
Versioning, calibration and corrections
Reports identify their methodology reference and status, snapshot date and later review date. A public aggregate methodology version is included only after one has been assigned and approved; independently versioned components do not constitute that aggregate version. Material interpretation changes are recorded in correction history.
Scoring
The current Dutch furniture report publishes no numeric Adoption Score. Company-capability states are presented directly rather than converted into a public ranking.