ReviewDesk guide · By the Ingenious AI team · Published 14 November 2026 · Reviewed 14 November 2026
What does review coverage mean — and what breaks when requests go unanswered?
Generate a free adverse-outcome letter — no sign-up Build a free pre-use notice See pricing
The short answer. Coverage is the ratio of answered requests to received requests, measured from real cases — ReviewDesk's own counter reports answered-over-total from the desk's actual records and shows an honest empty state when nothing has been routed yet. Unanswered requests break three things at once: the legal promise the notice made, the operational record the audit trail needs, and the person's willingness to believe anything else you say. Coverage is therefore the programme's vital sign.
What is review coverage?
Review coverage is the proportion of human-review requests (and their siblings — corrections, opt-outs, appeals) that receive a decided, communicated outcome, measured over the real case volume. In ReviewDesk the concept is embodied in the desk's response-coverage gauge, which reads the workspace's own answered-over-total from actual cases — never a sample figure — and the inbox discipline exists to keep that coverage complete.
Why is coverage the number that matters most?
Because it is the only metric the person on the other end experiences directly. Your notice quality, your letter craft and your trace discipline are all invisible to the applicant whose request vanishes; coverage is the programme as lived. It is also the metric that cannot be flattered by selection: volume is whatever arrived, answered is whatever the docket shows, and the ratio is arithmetic. Programmes drift into reporting activity — letters sent, notices published — because activity is controllable. Coverage reports the promise kept, which is why it belongs on the leadership panel ahead of everything else.
ReviewDesk's landing page makes the same commitment in public: the coverage gauge reads the workspace's real answered-over-total and sits empty, honestly labelled, until cases exist. That design — show nothing rather than a sample — is the 'not recorded' ethic from the evidence guide applied to marketing. If your internal reporting would embarrass you on a public page, the reporting is the problem.
What breaks legally when a request goes unanswered?
The promise chain. The notice offered a route; the route was used; the process then failed to perform the act the entire framework exists to guarantee — a person's examination of the case. In Colorado terms the explanation and review machinery is the duty, not a courtesy appended to it, and an unanswered request is the cleanest possible evidence that the machinery is decorative. It also corrupts the surrounding record: the audit trail shows a request with no outcome, which reads worse than no request at all, because it proves the organisation knew and did not act.
Prioritise accordingly. If capacity forces triage, the statutory-clock cases and the oldest cases outrank everything, and a request answered late with candour — the escalation pattern in the escalation guide — is a recoverable position. A request never answered is not. Coverage debt compounds silently because each unanswered case remains technically answerable until the person's situation makes the answer worthless: the home let, the role filled, the renewal lapsed.
What breaks operationally?
The feedback loops all close through answered cases. Corrections reveal data quality only when someone verifies them; overturn patterns reveal criteria problems only when reviews conclude; reviewer calibration exists only across decided cases. An inbox with a growing unanswered tail is a sensor network being unplugged one node at a time, while the dashboards built from the answered minority grow steadily less representative. Selection bias enters exactly where the programme congratulates itself on its overturn rate — computed, unnoticed, over only the cases someone reached.
The tail also hides its own contents. Unanswered requests are not a random sample; they skew toward the hard, the ambiguous and the externally blocked — precisely the cases carrying the most information and the most risk. The coverage discipline — one docket, visible ageing, escalation computed on read — exists to fight that skew mechanically, since goodwill alone demonstrably does not. The design is in the inbox guide.
What breaks reputationally?
Belief. An unanswered request teaches its author one lesson — this organisation's stated process is theatre — and people generalise lessons like that across everything the organisation says afterwards: its notices, its marketing, its job adverts. The reputational damage is disproportionate to the administrative failure because it is evidence-based from the victim's side; they are not repeating a rumour, they are reporting an experiment they personally ran on your process and watched fail. In review-and-ratings terms, the unanswered are also the motivated: a person with a story about silence writes it where future applicants will read it.
Contrast the answered adverse outcome. People routinely accept decisions they dislike from processes that answered them — the letter engaged, the review reasoned, the timeline honoured. Acceptance is purchasable with responsiveness at a price any operating budget can afford, which is what makes the unanswered tail such an own goal: the cheapest trust in the entire programme, left unclaimed in a queue nobody swept.
How do you measure coverage honestly?
Count from the docket, in the open: requests received in the period, requests decided and communicated, median days to outcome, age of the oldest open case, breaches with causes. Answered means both halves — outcome recorded and reply sent — the closure discipline from the inbox guide. Coverage computed any other way (surveys, sampling, the owner's impression) will flatter the programme, because the unanswered are by definition the cases nobody is looking at.
Publish the panel internally on a rhythm and pair coverage with its quality twin — the outcome-note sampling from the history guide — because coverage achieved by rubber-stamping is worse than honest backlog: it converts a visible operational problem into an invisible integrity problem. Speed and substance are the two dials; the panel watches both or it watches neither honestly.
How do you recover coverage once it has slipped?
Declare the backlog, staff it, and answer oldest-first with candour about lateness. The worst recovery strategy is silent catch-up that pretends the timeline was normal; the record shows the stamps, and applicants compare notes. A short communication to the waiting — your case is with us, here is the honest position, here is your status link — resets expectations at near-zero cost and stops the complaint inflow that backlog generates. Then fix the cause, which the backlog's composition will name: capacity, routing, a broken form, a decision type whose volume outgrew its bench.
Treat recovered coverage as a programme event, minuted with its cause analysis, and let the risk assessment carry the story — a period of slipped coverage, its cause and its remedy is exactly the kind of honest self-knowledge an assessment exists to record, per the assessment guide. Programmes are not judged, by fair examiners, on never slipping. They are judged on whether the slip was seen, owned and engineered against. Coverage is how you prove all three.
Your five actions for review coverage
- Measure coverage from the docket only: received, decided-and-communicated, median days, oldest open case, breaches with causes.
- Treat an unanswered request as the programme's worst state — worse than a late answer, which candour can still redeem.
- Fight the tail's skew mechanically: one docket, visible ageing, escalation computed on read.
- Pair coverage with outcome-note quality sampling so speed never quietly eats substance.
- Recover slipped coverage oldest-first, with an honest update to everyone waiting and the cause fixed and minuted.
Worked example: the eleven per cent nobody was counting
A fictional services group — Meridian Support — believes its review handling is strong: letters praised, reviewers conscientious. Its first honest coverage count, run from the newly unified inbox, says otherwise: of 214 requests in the quarter, 24 have no outcome — eleven per cent, concentrated in one decision type whose requests arrived through a legacy form that filed into a folder nobody owned. The reviewers were indeed conscientious — about the cases they could see. The system's failure was routing, and it was invisible precisely because every visible case was handled well.
Recovery takes a fortnight: the folder is drained into the docket, the waiting applicants receive honest updates with status links, and the oldest cases are decided first — three outcomes change, including one person's job offer restored for the next intake. The legacy form is retired the same week. Meridian's leadership keeps the eleven per cent in the programme's history, unedited: it is the number that taught them coverage is counted from arrivals, never from the cases that happen to reach a desk.
Frequently asked questions
What is review coverage?
The share of requests received that are answered with a decided, communicated outcome, counted from the real docket. ReviewDesk's own coverage gauge reads the workspace's actual answered-over-total — and shows an honest empty state before any cases exist.
Why is an unanswered request worse than a late one?
A late answer can still remedy the case and candour can redeem the timeline. An unanswered request breaks the notice's promise entirely, leaves a knowing gap in the audit trail, and teaches the person that the stated process is theatre.
How should coverage be measured?
From the docket alone: received, decided-and-communicated, median days, oldest open case, breaches with causes — paired with outcome-note quality sampling so coverage is never bought with rubber stamps.
What usually causes coverage to slip?
Routing failures (requests landing outside the owned queue), capacity outgrown by one decision type, and shadow channels where staff answer off-docket. The backlog's composition names the cause; fix that, not the symptom.
Is this guide legal advice?
No. ReviewDesk is templates and workflow tooling, not legal advice. Coverage is described as ReviewDesk's operating discipline; consult counsel for your obligations.
Keep reading: How do you answer every human-review request — without sounding like a bot? · How do you run ADMT workflows across Colorado, California and Illinois at once? · How can an agency run ADMT desks for multiple client businesses? · How can HR-tech resellers offer ADMT notice and review tooling to their clients? · How can a small business meet ADMT notice and review duties without a compliance department? · All ReviewDesk guides · ReviewDesk home · Try the free desk · Free letter generator · Free notice builder · Pricing
ReviewDesk is built by Ingenious AI. Templates and workflow tools — not legal advice.