Ranking

A shortlist you can defend

Mandatory filters, taxonomy match, semantic retrieval, evidence assessment and weighted scoring stay five separate layers — so relevance is never mistaken for qualification.

Overview

Explainable shortlist

When a supplier selection is challenged, the reasoning usually lives in a specialist’s head, which is exactly where it cannot be inspected. SpiderNet keeps matching in five distinct layers and shows the result as a requirement-by-requirement matrix with evidence counts, gaps, contradictions and risk flags, so the shortlist can be opened up rather than asserted.

What it does

Every score opens onto evidence

Five separate layers: mandatory filters, taxonomy, semantic retrieval, evidence assessment, weighted scoring.

A requirement-by-requirement matrix per supplier.

Evidence counts, gaps, contradictions and risk flags shown.

Pool origin is visible — approved, known potential or external.

An approved shortlist is immutable.

How it works

How the shortlist is produced

1Confirmed mandatory criteria filter the pool
2Taxonomy and semantic layers assess fit
3Evidence is assessed requirement by requirement
4You review, adjust with reasons, and approve
FAQ

Common questions

Can we add or remove a supplier by hand?

Yes, with a recorded reason — then the approved shortlist becomes immutable.

Can a supplier pay for better placement?

No. Ranking neutrality is an architectural constraint: supplier-side monetisation never affects qualification, matching or ranking.

Is semantic similarity the same as qualification?

No, and keeping them separate is the point — a relevant-looking supplier can still fail a mandatory requirement.

Explore more

More of what SpiderNet does

Get started

See one of your own SOWs shortlisted

Request a demo to turn a real statement of work into an evidence-backed supplier shortlist.