The twelve
"Fleet management" rose 30% month on month to August 2026 — a rare double-digit rise for a head term. What owners are searching for is not the definition. It is a way to tell whether their fleet is being run well.
Here are twelve indicators, each with the definition to insist on and the way it gets gamed. The definition matters more than the number; almost every dispute between owner and manager over performance turns out, on inspection, to be a disagreement about arithmetic.
1. Unplanned technical off-hire days
Definition: days the vessel was unavailable to the charterer due to technical cause, excluding scheduled dry docking.
Target: under 3 days per vessel per year for a well-maintained mid-age ship.
Gamed by: reclassifying technical off-hire as "owner's convenience" or folding it into a docking window.
2. Planned maintenance compliance
Definition: critical jobs completed within their due window ÷ critical jobs due. Critical only — total-job compliance hides the ones that matter.
Target: above 95% on critical; above 90% overall.
Gamed by: re-baselining due dates, or classifying jobs as non-critical.
3. OPEX variance by cost head
Definition: actual vs approved budget, by head, with prior-year comparison.
Target: within ±5% at total level; explained at head level.
Gamed by: deferring maintenance into the next budget year. Always read variance alongside overdue PMS jobs — a favourable repair variance with rising overdue jobs is not a saving, it is a loan.
4. Officer retention
Definition: officers returning for a further contract ÷ officers eligible to return.
Target: above 85% for a stable fleet; senior ranks reported separately.
Gamed by: using total contracts signed as the denominator, which flatters heavily.
5. Relief compliance
Definition: reliefs completed within contract end date + agreed tolerance ÷ total reliefs due.
Target: above 95%.
Why it matters: overdue reliefs are a fatigue risk, an MLC exposure and the leading indicator of retention decline.
6. PSC deficiencies per inspection
Definition: total deficiencies ÷ inspections, over a rolling 24 months.
Benchmark: the Paris MoU recorded 51,797 deficiencies across 16,474 inspections in 2025 — roughly 3.1 per inspection across all vessels inspected — with a 4.18% detention rate.
Gamed by: counting only inspections that produced findings.
7. Lost Time Injury Frequency (LTIF)
Definition: lost time injuries × 1,000,000 ÷ exposure hours. State the exposure hour convention.
Gamed by: reclassifying lost-time cases as restricted-work cases. Always read LTIF next to near-miss reporting rate: a falling LTIF with a falling near-miss rate usually means reporting has stopped, not that safety has improved.
8. Near-miss reporting rate
Definition: reported near misses per vessel per month.
Target: rising is good. A fleet reporting fewer than one per vessel per month has a reporting culture problem, not a safety record.
9. CII rating and trajectory
Definition: current-year attained CII and projected year-end rating per vessel.
Why: ratings tighten annually. A vessel at C today is at D tomorrow on the same operating profile.
Gamed by: reporting the prior year's confirmed rating rather than the live projection.
10. Hull and propeller performance
Definition: speed loss (or added power) versus a reference condition, corrected for weather and draft.
Target: under 5% speed loss between cleanings.
Why: it is the cheapest fuel and emissions lever available, and it decays continuously.
11. Critical spares stockout rate
Definition: occasions a critical spare was unavailable when required ÷ critical spare demands.
Target: near zero.
Gamed by: narrowing the definition of "critical".
12. Fuel consumption vs charter party warranty
Definition: actual consumption vs warranted, in good weather as defined in the charterparty.
Why: it is the source of most performance claims, and it links technical condition directly to commercial exposure.
The dashboard rule
| Layer | Audience | Frequency |
|---|---|---|
| 4 headline KPIs | Owner / board | Monthly |
| 12 KPIs above | Fleet manager | Monthly |
| Vessel detail with drill-down | Superintendent | Continuous |
If the owner's four cannot be traced to the superintendent's detail in two clicks, the reporting chain is broken somewhere.
What not to track
- Cost per day in isolation. Meaningless across different flags, crew nationalities and docking positions.
- Total PMS jobs completed. Volume without criticality is noise.
- Audit findings closed. Without ageing and recurrence, closure is administrative.
- Fuel consumed. Without normalisation for weather, draft and speed, this is a weather report.
benchmark figures derived from the Paris MoU Annual Report 2025 (16,474 inspections, 9,879 with deficiencies, 51,797 deficiencies, 688 detentions). Targets are indicative industry practice. Reviewed by the Zeaclub Editorial Team, 24 August 2026.
Frequently asked questions
How many KPIs should a fleet report on?
Twelve at fleet-manager level, four at owner level. Beyond that, attention dilutes and nothing gets acted on.
Should KPIs be in the management contract?
Increasingly yes, with definitions attached. Avoid bonus-malus structures on any KPI that the manager also calculates — measure first, incentivise later, and only where the definition is unambiguous.
What is a good detention rate?
Zero is the only acceptable target, but context matters: the Paris MoU fleet-wide rate was 4.18% in 2025. What distinguishes a good manager is what changed after the last detention.
How do I stop KPIs being gamed?
Publish the definitions, insist on the raw denominators, and always read paired indicators together — LTIF with near-miss rate, cost variance with overdue maintenance, retention with relief compliance.