Compute billable utilization per technician for a period — billable hours logged over
available capacity — against role-appropriate targets. This is a leadership economics
instrument for pricing, hiring, and workload decisions. (Billable Analysis slices the billable
hours themselves; this skill divides them by capacity and compares to targets.)
1. Confirm the period (default: last full month), the roster (look up the members), each
person's capacity baseline (default 40h/week minus known PTO — ask; do not guess PTO), and
the role-aware targets. If leadership has no targets, propose the conventional bands and
label them conventions, not benchmarks: dedicated reactive tech 70-80%, escalation/senior
50-65%, lead with management duties 30-50%, dispatcher/coordinator near 0% by design.
2. Read the time entries per tech for the period, split billable vs non-billable as this desk
marks them. Record caps — capped pulls understate utilization; split by week or client
until uncapped, or report floors.
3. Compute per tech: billable hours, total logged hours, capacity, billable utilization, and
logging coverage (total logged / capacity). LOW utilization with LOW logging coverage is a
time-entry compliance question, not a utilization finding — route those rows to time-entry
compliance before anyone reads them as idle time.
4. Compare each tech to their ROLE's target band, never to the desk average. Flag over-target
as sustainability risk (a tech at 95% has no slack for escalations, training, or a bad week
— that is a burnout and quality flag, not a gold star).
5. Aggregate view: blended utilization, trend vs prior period if available, and the capacity
headroom answer in hours ("roughly N billable hours/month of absorbable capacity,
concentrated in <roles>").
6. Output: per-tech table (role / capacity / billable / utilization / role target / coverage
flag), the sustainability and compliance flags, the aggregate read, and a methodology note
— capacity assumptions, billable definition used, targets source, searches run, caps hit.
Guardrails: LEADERSHIP-ONLY — this report prices people's time; it must not circulate as a team
leaderboard. Say so in the output header. Utilization ≠ worth, stated in the report itself:
mentoring, documentation, automation-building, and presales are low-billable and high-value; a
utilization number captures none of it. Never rank techs by utilization alone, never volume
alone. Role-aware targets are mandatory — comparing a dispatcher or escalation engineer to a
reactive-bench target is a category error this report must refuse to make. Low numbers get the
compliance check before the idle interpretation, every time. Never present capped time totals
as complete; disclose caps and direction of bias. PTO and part-time capacity come from the
requester — never assumed silently.
Magic Library