Demand
Co-specification edges computed from 36 months of real project activity — which materials are sampled together, considered together on boards, and how strongly, expressed as lift against chance. Aggregated to counts at the project grain, never to a member or customer identifier.
Gray is ceding share.
recent four quarters vs the four before
23,496,595
Spec lines observed
211,816
Materials tracked
A pairing rail serves three kinds of evidence, in order — and the three tiers are never blended into one claim.
Order evidence
Sampled or specified together on the same real project — the strongest signal, and the only one presented as demand-lift co-occurrence.
Board activity
Considered together on boards — shown only when sampled order evidence isn't available for that material.
Colour top-up
A look-alike by palette similarity — the fallback tier, and never presented as specified together.
Honesty
Coverage is partial — about 39% of materials carry order-derived companions today, because order history doesn’t touch the whole catalog. The number is stated rather than smoothed over.
A material without co-specification data shows nothing rather than a guess — and a colour look-alike is labelled as one, never presented as though it were order evidence.
For machines
Ask for a material’s pairings and the rail comes back tiered — order evidence first, board activity only when order evidence is missing, colour last, each item labelled with its source.
Explore the developer surface →curl "https://beta.materialgraph.com/api/v1/materials/6g23prj45moq6scfh2mn4/pairings" \
-H "Authorization: Bearer mg_read_..."