ReviewDesk
← Back

ReviewDesk guide · By the Ingenious AI team · Published 14 November 2026 · Reviewed 14 November 2026

How do SLA clocks keep human reviews inside their deadlines?

Generate a free adverse-outcome letter — no sign-up Build a free pre-use notice See pricing

The short answer. A deadline becomes an SLA clock when the case itself carries it: the date it counts from, the rule it counts under, the days remaining, computed whenever the case is read. ReviewDesk's paid plans run statutory clocks — the Colorado 30-day countdown with overdue stamps — and let you set tighter internal targets per workflow, so 'when is this due?' is a property of the case, never a question for a spreadsheet.

What is human-review SLA clocks?

An SLA clock is the deadline machinery attached to a review case: a due date derived from a stated rule and a stated start event (deadline with provenance), a countdown computed on read rather than in periodic reports, and escalation states that surface as the deadline approaches or passes. In ReviewDesk the SLA workflow adds assignment, deadlines, escalation and applicant status to the human-review inbox on paid plans.

Why do deadlines fail without clocks?

Because in most organisations a deadline is stored in a different place from the work. The decision sits in the screening platform, the date sits in a spreadsheet, the request sits in an inbox, and the person who could join them up is on holiday. Each artefact is accurate; the system fails anyway, at the joins. A clock attached to the case removes the joins: the case is created with its start event, the rule computes the due date, and every view of the case shows the remaining days. Nobody can look at the work without seeing the time.

The second failure is report latency. A weekly overdue report tells you on Friday about the case that expired on Tuesday. Computing urgency on read — each time the case is opened or listed — means the board is never stale, and escalation states are facts, not findings. ReviewDesk's Phase-8 workflow computes escalation exactly this way: on read, from the case's own dates, emitted once. The design principle generalises beyond this product: any deadline that matters should be computable from the record it governs, at the moment anyone looks.

What is a deadline with provenance?

A due date that can explain itself: this case is due on 14 April because the decision was made on 15 March and the Colorado explanation window is 30 days — the start event, the rule, the arithmetic, all visible on the case. Provenance kills the two commonest deadline disputes: 'who set this date?' and 'from when does it run?'. It also protects the team: when an applicant challenges a timeline, the answer is a reading, not an investigation. ReviewDesk records the provenance with the deadline, so the case file and the clock are one artefact.

Provenance matters equally for internal targets. An acknowledgement target of two working days set by your own policy is a promise you made yourselves; the case should say so, distinguishing it from statutory clocks, because the two kinds of lateness have different consequences and different fixes. Conflating them trains the team to treat all clocks as approximate — the first step toward treating the statutory ones that way too.

How should escalation be staged?

In three visible states before the breach, not one alarm after it. Approaching: the case enters its final third and the board marks it. Due soon: the assignee and the inbox owner both see it, and reassignment is one click if the assignee is stuck. Overdue: the case stamps itself, surfaces to leadership views, and the record keeps the overdue fact permanently — late is data, and hiding it helps nobody. ReviewDesk stamps overdue cases rather than quietly extending them; an extended deadline is a new deadline with its own provenance entry explaining the change.

Escalation should also have a human content, not just a colour. The escalation note answers: what is blocking this case — evidence awaited, reviewer capacity, a third-party record — and what happens next. Applicants experience escalation through the status link, per the status guide; a case that visibly escalates with a reason reads as supervised, while a case that simply ages reads as abandoned, even when the work behind both is identical.

Which clocks should a workflow run?

Start with the statutory map: Colorado's 30-day explanation window where it applies; your regimes' other dated duties as they come into force. Add internal clocks where the person's situation imposes its own urgency — hiring reviews inside the live requisition, letting corrections inside the availability window, triage disputes inside days. Then add one acknowledgement clock for everything: a fast, cheap promise — two working days is a common shape — that guarantees no request ages in silence even while its substance is worked on.

Resist clock inflation. Ten targets per case teaches the team to ignore nine. Three clocks — acknowledgement, internal decision target, statutory backstop — cover most workflows, in ascending order of consequence. The art is in the gaps between them: internal targets should bite well before statutory ones, so a statutory breach requires two earlier failures, each visible, each owned. That is what a control is: not a date, but a sequence of opportunities to notice.

How do clocks interact with case assignment?

A clock without an owner is a poster. Assignment binds each case to a named reviewer whose own queue shows their clocks in order, and the inbox owner's board shows every clock across the team. Reassignment moves the whole case — file, clock, provenance — to the new owner as a recorded event, so absence and workload spikes are handled by moving cases, not by hoping. The mechanics are in the assignment guide; the clocks are what make assignment urgent rather than administrative.

Measure reviewers by quality-weighted timeliness, never by raw closures. A league table of closures teaches reviewers to close easy cases and stamp thin notes; a measure that pairs on-time completion with outcome-note sampling — against the review standard — teaches the tempo you actually want: brisk, substantive, honest. Clocks set the tempo; the standard sets the substance; confusing the two produces fast, empty reviews at scale.

What do you do when a clock is breached?

Record, remedy, learn — in that order. The breach stays on the case permanently; the remedy for the person is completed as fast as possible with the lateness acknowledged plainly in the communication, because applicants forgive admitted lateness more readily than they forgive discovering it. Then the learning step: classify the breach's cause — capacity, evidence wait, assignment failure, clock misconfiguration — and count causes monthly. Breaches with a cause distribution are a management instrument; breaches treated as individual failings are a morale instrument pointed the wrong way.

Systemic causes get systemic fixes: a second reviewer for a decision type, an evidence-chasing template, a clock corrected at the workflow level. The annual risk assessment consumes the breach statistics as evidence about the programme's safeguards — see the assessment guide. A programme with a small, analysed, shrinking breach count is demonstrably supervised; a programme claiming zero breaches across thousands of cases is claiming something its own inbox would not corroborate.

Your five actions for human-review SLA clocks

  1. Attach each deadline to its case with provenance: start event, rule, arithmetic — visible wherever the case is viewed.
  2. Compute urgency on read, not in periodic reports, and stamp overdue cases permanently rather than quietly extending them.
  3. Stage escalation in visible steps with a human note: what blocks the case and what happens next.
  4. Run three clocks per workflow — acknowledgement, internal decision target, statutory backstop — with real gaps between them.
  5. Bind every clock to a named assignee, and measure reviewers on timeliness paired with outcome-note quality.

Worked example: thirty days, watched properly

A fictional Colorado employer — FrontRange Clinics Group — issues about 300 adverse screening outcomes a month. Its explanation cases carry the 30-day statutory clock with provenance, an internal target of ten days, and a two-day acknowledgement clock. The board's morning state is a single glance: 41 open explanations, median age four days, two cases in their final third — both already drafted, awaiting the applicants' correction documents, with escalation notes saying exactly that.

Across the quarter, one case breaches: a third-party record check stalls, the escalation fires, the case is reassigned and completed on day thirty-three with the lateness acknowledged in the letter and the cause logged as evidence-wait. The quarter's statistics — one breach, cause known, remedy recorded — go into the programme review as they are. Nobody at FrontRange describes the quarter as perfect. They describe it as watched, which is the quality the clocks were bought to produce.

Frequently asked questions

What is an SLA clock in human review?

Deadline machinery on the case itself: a due date derived from a stated rule and start event, days remaining computed whenever the case is read, and escalation states as the date approaches. ReviewDesk's paid plans run the Colorado 30-day statutory clock this way.

What does 'deadline with provenance' mean?

The case shows why its deadline is what it is — the decision date it counts from and the rule it counts under — so the timeline can be explained to the applicant or an auditor by reading the case, not by investigating it.

Should internal targets match the statutory window?

No — they should bite well before it. An acknowledgement clock in days and an internal decision target comfortably inside the statutory window mean a statutory breach requires two earlier, visible failures first.

What happens when a case goes overdue?

It stamps itself overdue permanently on the record, escalates visibly with a note on what blocks it, and is completed with the lateness acknowledged. Breach causes are counted monthly so systemic fixes follow.

Is this guide legal advice?

No. ReviewDesk is templates and workflow tooling, not legal advice. The clocks are described as ReviewDesk implements them; consult counsel for your deadlines.

Keep reading: What should happen when a human-review request goes overdue? · How does an applicant status link change the experience of waiting for a review? · How should human-review cases be assigned across a team? · What should a review case's decision history record? · How does the 3-year evidence vault export a complete decision file? · 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.