A method built to be audited.
We measure the digital shelf continuously and the physical shelf in the store. The same five steps apply to both. Every step is documented.
Observation
Digital storefronts are measured continuously. Physical fixtures are observed in the store, under a trained field protocol, with every visit timestamped and geolocated.
Systematic reading
Every observation is converted to structured data: what was stocked, how it was faced and positioned, and what the tag beside it said.
Catalog reconciliation
Observations are matched against a normalized catalog of 13,700+ products, so an item on the shelf and its digital listing become one fact.
Independent audit
Separate verification passes challenge what the first reading claimed before anything is published. Findings that fail audit are reverted, and the reversal is logged rather than erased.
Versioned delivery
Facts land in a versioned warehouse with provenance on every row, including every correction since first observation. Delivery is by client dashboard, spreadsheet-ready extract (CSV or Excel), warehouse share, or the API, on the cadence the engagement requires. Clients can ask why a number is what it is, and get an answer.
Every observation carries a timestamp, a store, and a photograph.
That is the whole trust model. Not a brand promise, a paper trail.
| Products under measurement | 13,700+ |
| Brands tracked | 2,900+ |
| Categories in the record | 1,600+ |
| Price observations | 46,000+ |
| Physical facings measured | 970+ |
| Market | one major US metro market (named under license) |
| Coverage weighting | independents and specialty first |
| Digital-shelf cadence | continuous |
| Physical capture cadence | per-visit, by banner |
Figures are live warehouse counts, restated as floors. Updated with each program release.
A measurement firm that hides its limits is selling something else. Our physical panel is young and geographically concentrated. Facing counts are per-visit observations, not continuous telemetry. Any systematically read shelf contains residual error. That is why the audit passes exist, and why corrections are logged rather than overwritten.
Where the record is thin, we say so, in the data itself, row by row.
