Why Your Content Queue Keeps Growing While Nothing Ships

The first symptom is a widening start-to-finish gap

Track how many items entered active work and how many reached the agreed definition of done during the same period. A persistent excess of starts means inventory is accumulating somewhere, even if everyone is occupied. Pair that count with the age of the oldest item. A large queue of young cards after a planned intake may be benign; a smaller queue with several very old cards can signal blocked review, unclear ownership, or work that should be cancelled.

Do not diagnose from the calendar alone. Dates may move while cards remain nominally “scheduled,” and published pages can conceal skipped refreshes or quality debt. The Kanban Guide’s flow measures help separate quantity, elapsed time, and active inventory. They identify where to investigate, not why the problem exists. The next step is to compare observable traces for competing causes.

Evidence: Kanban Guides

Separate intake overload from downstream constraint

If new selections spike while review capacity stays stable, the likely mechanism is intake overload. If intake is stable but drafts wait longer in one column, a downstream constraint is more plausible. If cards bounce backward repeatedly, unclear entry criteria or mismatched quality expectations may be driving rework. If items remain active while their strategic reason has expired, weak cancellation governance is occupying capacity.

Build a cause table with four rows: overload, bottleneck, rework, and stale commitment. For each row, list the trace you expect, evidence that would contradict it, and one inexpensive check. Avoid attributing congestion to motivation until workflow evidence rules out structural causes. The Scrum Guide’s shared goal and completion definition provide useful questions: is the work focused, and do contributors agree what finished means?

• Overload trace: selections rise before age rises across several stages.
• Constraint trace: age clusters immediately before one scarce capability.
• Rework trace: cards cross the same boundary more than once.
• Stale-work trace: active cards no longer support a current reader or business decision.

Evidence: Kanban Guides; Scrum Guides

Watch maintenance displacement and batch behavior

A team can preserve its headline publishing rate by postponing updates, corrections, source checks, or accessibility work. Add a maintenance-age view so apparent throughput cannot hide library deterioration. Also inspect batch arrivals: if reviewers receive many drafts on one day, the calendar may be smoothing dates while the workflow creates periodic overload. Measure the interval between handoffs, not just the final publication date.

DORA’s balanced delivery perspective is relevant as an independent caution: speed should be read alongside stability and quality, and metrics should not become competitive targets divorced from context. Editorial indicators require local definitions. A return from review might represent healthy quality control or preventable ambiguity; distinguish the reason codes before treating the count as good or bad.

Evidence: Kanban Guides; DORA / Google Cloud

Use an aging-board diagnostic

Color each active card by age relative to other items of the same class: ordinary, watch, or investigate. Do not use one threshold for a correction and a research-heavy guide. On every investigate card, record last meaningful movement, current blocker, the person able to resolve it, and whether the item still deserves capacity. Then total cards by stage and age band. The visual should make queues and neglected work undeniable without naming individual performance scores.

Choose the oldest consequential card and walk its history. Verify whether it was actually ready when pulled, where evidence or decisions were missing, and what other work entered while it waited. That trace often reveals a policy problem faster than a general retrospective. If the item cannot be completed responsibly, cancelling it is a legitimate flow improvement; merely moving it to a hidden backlog is not.

Evidence: Kanban Guides; Scrum Guides; UK Government Digital Service

Distinguish a temporary shock from a broken policy

An absence, platform incident, major correction, or campaign can create a short queue without invalidating the operating model. Look for recovery after the shock ends and whether the expedite policy documented what was displaced. A policy problem persists across cycles or recurs under ordinary demand. Examine at least several replenishment periods before claiming a trend, while acting immediately on safety, rights, or factual-correction issues.

Illustrative example: eight drafts wait for one evidence reviewer after a campaign. The cause could be a temporary batch, chronic reviewer scarcity, or drafts arriving without source records. Checking submission completeness and queue age across prior weeks separates those explanations better than adding another writer. This scenario is constructed and does not report TenMultigure performance or prove that one observation establishes causality.

Evidence: DORA / Google Cloud; UK Government Digital Service

Select one repair and protect the observation window

For overload, tighten replenishment and stop opening new work. For a constraint, redirect capacity, reduce batch size, or change the service policy. For rework, improve the upstream exit rule and sample a few returned cards. For stale commitments, add expiry and cancellation authority. Change only the most plausible policy, state the expected trace, and observe two or more cycles before layering another intervention.

The diagnostic is successful when the explanation narrows and the oldest important work moves or closes—not when the board turns green. Flow statistics cannot establish reader value, and a faster system can publish weak content more efficiently. Pair the repair with evidence and quality checks. Review this diagnosis earlier than 2027-02-10 if staffing, editorial standards, or content mix changes materially.

Evidence: Kanban Guides; DORA / Google Cloud