Contact data decays continuously. Messages fail silently. And the losses – from unfilled vacancies to unpaid bills to missed appointments – are scattered across budgets where nobody sees the total. We propose a framework to change that.
Digital communication has been built almost entirely around endpoints – email addresses, mobile numbers, postal addresses, messaging handles. An endpoint is treated as if it were the person. It is not.
Endpoints are inherently perishable. People change jobs and lose work emails. They switch networks and abandon numbers. They move home. They migrate between platforms. Business contact data decays at roughly 2% per month – around 22.5% per year (Prospeo, Apollo, ZeroBounce, 2025–2026).
The result: every organisation running a large contact database is fighting a war against entropy it cannot win. It can slow the decay – with validation tools, data enrichment, CRM cleansing – but none of these arrests it. They slow the treadmill. They do not get the organisation off it.
The organisation’s contact database is not a stable record of who its customers are. It is a perishable snapshot of where they were when last seen.
We propose CRF as a distinct, measurable KPI – one that every organisation already pays, in two categories, but that almost none of them isolates or manages.
The maintenance costs sit in IT and data-governance budgets. The failure costs are dispersed across recruitment, collections, customer service and compliance. No single owner sees the total, so no one manages it. CRF is the metric that makes the total visible.
CRF is not hypothetical. It surfaces in publicly available figures from every communication-intensive sector – as bookable losses currently mis-attributed to “operations” rather than to reachability.
Global corporate hiring costs per year. Derived from the SHRM 2025 benchmark of USD 5,475 per non-executive hire, applied to an estimated 120–150 million formal corporate hires annually (ILO labour data; Staffing Industry Analysts Global Staffing Report, 2024). Consistent with SIA’s estimate of ~USD 650 billion in global staffing industry revenue. Reachability failure – candidates unreachable when vacancies open or redeployment windows close – represents a currently unmeasured fraction of this total.
Average cost per hire (non-executive), 44-day time-to-fill. Every unreachable candidate extends the vacancy and multiplies this cost. SHRM, 2025.
Estimated OECD domestic energy arrears. Ofgem reports 8.4% of UK domestic customers in arrears, averaging ~£1,815 per account. Applied to approximately 450 million formally billed OECD households, and discounted for markets without the UK’s 2022–2023 energy-price shock, this produces an OECD-wide working estimate of USD 60–100 billion, midpointed at ~USD 75 billion. Significant uncertainty applies; this is a working estimate.
UK domestic energy debt, June 2025, up 71% since 2023. Nearly three-quarters held by customers with no repayment plan – predominantly unreachable ones. Ofgem, 2025.
Global rail delay-compensation costs per year. DB’s €197M across ~2 billion journeys implies ~€0.10 per journey. EU rail carries ~9 billion journeys per year (Eurostat, 2024); UK Delay Repay handles ~2.6 million claims per quarter (ORR, 2024–25). Adding conservative estimates for Japan, South Korea and Australia – markets with comparable compensation regimes – gives a working estimate of €1–2 billion globally, midpointed at €1.5 billion. DB is an outlier for punctuality; global rates are adjusted downward accordingly.
Deutsche Bahn passenger compensation, 2024; 6.9 million claims. Every claim pre-emptable by reaching the passenger in time is a direct expression of CRF. DB / dpa, 2025.
Estimated OECD missed-appointment costs per year. NHS England data implies a 1.9% primary-care DNA rate (£30/appointment). Esendex analysis (2024) estimates 11.8 million hospital DNAs in the UK at ~£1.9 billion. Scaling both proportionally to OECD populations (1.2 billion vs. UK’s 67 million) at a conservative USD 40 average appointment cost produces a working OECD estimate of USD 20–30 billion. DNA rates in lower-income systems are typically higher, suggesting the true global figure is materially larger.
Annual cost of missed GP appointments, UK; 7.2 million per year. Most common reason: patient was unaware, forgot, or received incorrect information. NHS England, 2019.
The full research report also covers banking, insurance, logistics, debt collection, telecoms and further sectors. All figures are cited with precise descriptions of what each represents. RELAY estimates are explicitly labelled.
Every solution the market has offered – better CRM, AI-driven enrichment, phone-number verification APIs, real-time data-quality platforms – is endpoint-centric. They try to keep pace with the decay of stored contact details. In a race against continuous, exogenous entropy, they can at best slow the rate of loss.
They do not stop it, because the decay is not caused by the organisation’s negligence. It is caused by the customer’s life. The only structural solution is to stop addressing endpoints and start addressing people.
This is the same move the internet has already made twice.
Email made the same move – a portable address that survived ISP changes. But email’s abstraction was only partial: the email address itself is now perishable, tied to an employer or platform the person will eventually leave.
A domain name points to a server. A relay address points to a person.
Every person, organisation and AI agent receives a persistent relay address – for example name.relay – as a single, durable communication endpoint. The sender needs to know only the identity; RELAY routes the message to wherever and however the recipient currently chooses to receive it.
For an organisation, this collapses the CRF maintenance category: there are no endpoints to store, validate or cleanse, because the routing is maintained by the recipient. And it compresses the failure category: messages reach a live, currently active channel rather than a stale number or an abandoned inbox.
The relay address is not an app, an inbox or a messaging service. It is an infrastructure layer – the same kind of abstraction DNS provides for the internet, now applied to the person.
Available as an accessible public whitepaper (11 pages) and a full research report for investors and enterprise stakeholders.