UK defence suppliers are operating in a 1% signal environment

The number

Over a two-week cycle, Quorion Signal processed 80,086 procurement events from the UK's core public procurement portals: Find a Tender Service, Contracts Finder, and the Defence Sourcing Portal.

Just 834 were relevant to defence and adjacent industrial supply chains.

That is 1.04% signal density. For every 96 notices published, one is relevant.

That is the operating reality for most UK defence suppliers. They do not lack access to procurement data. They operate inside a publication environment in which relevance is exceptionally scarce.

The noise field

Public procurement portals are publication systems, not intelligence systems. They are built to disclose notices, not to compress decision-making for commercial operators.

Across Quorion Signal's wider 16-channel ingest network during the same cycle, 102,819 events entered the system. After a seven-stage classification pipeline — source trust assessment, structural validation, noise filtering, canonical deduplication, node relevance screening, commercial relevance testing, and candidate routing — 2,663 survived. 100,156 did not.

Most of the raw flow died for structural reasons, not because the market was thin.

Looked at against total raw flow, roughly two thirds of all intake was off-node and a further 18% was cross-portal duplication. Together, those two failure modes removed about 85% of raw volume before any commercial judgement was required.

False abundance

A BD director opening Find a Tender or Contracts Finder sees abundance: hundreds of notices, constant publication, apparent activity. The data says something else. For a defence supplier, the overwhelming majority of that flow is either irrelevant, duplicated, non-biddable, or commercially late.

The interface makes this worse by flattening fundamentally different signal types into a single feed. In the classified universe for this cycle, Quorion identified:

Government intent is more visible in the public record than live bid-ready demand, yet a human scanner encounters both as undifferentiated notices.

By the time the formal ITT arrives, the more decisive work — problem definition, market testing, stakeholder alignment, competitor positioning — has often already happened.

Where the edge sits

The useful market is concentrated.

Across the wider classified universe, commercially meaningful activity concentrates around a relatively small number of buyers and buyer clusters. Passive horizontal monitoring is a weak posture. Buyer-specific pattern recognition is stronger.

Nuclear cluster: UKAEA, Sellafield, NDA and ONR combined.

The commercial implication

Most UK defence and industrial suppliers still operate a manual procurement monitoring model: senior commercial staff checking portals, forwarding notices, and layering keyword alerts over public feeds.

In a 1% signal environment, that model is structurally mis-specified.

The direct labour cost is real. At two hours a day, manual scanning consumes roughly £22,000 to £33,000 per year of senior BD time on a 220-day basis at £400–£600 per day, before overhead and before opportunity cost. But the larger cost is strategic. Firms arrive late. They mistake duplicated publication for market breadth. They over-weight formal tenders and under-weight pre-market movement. They spend scarce commercial attention where the market is already ossified.

Alerting does not solve that problem. Faster delivery of the same low-density feed simply accelerates waste.

The winning architecture is classification first: reduce the universe, separate live procurement from policy and pre-market movement, deduplicate across the portal estate, score against vertical and buyer fit, and route only the small number of signals that matter for a specific supplier position.

That is what Quorion Signal is built to do. Not tender alerts. Decision-grade classification, scoring and routing across regulated industrial markets.