for the dispatcher
The office's volume, efficiency, fairness
Calls, days, members. Office turnaround, crew confirmation, override rate. Days distribution, OOW respect, grievance count. Baseline comparison vs pre-Dispatch.
See the dispatcher view
positioning · how dispatch measures value
How Dispatch measures value.
Days dispatched, office turnaround, override rate, baseline comparisons. Here’s how each number gets computed, what it’s measured against, and why it tells the truth.
the unit problem
Days, not hours.
Live entertainment dispatch is measured in days and start times. Hours fluctuate — calls run the CBA minimum (5 hours for extras, 8 for heads) or stretch into 14-hour blowouts depending on what happens on the floor. Hours matter for the paycheck. Days matter for the workload.
Most software flattens this. We don’t.
Workload
Days on call
A 6-person, 3-day install = 18 days dispatched.
Volume
Calls dispatched
The 3-day install above = 1 call.
Rhythm
Start times
5am load-in vs midday corporate vs late-night strike.
Hours stay where they belong: on the per-day record, for payroll, OT, premium math, and xHand. They’re never the headline.
metric definitions
What each number means.
Operate-style definitions — no marketing copy. Every metric the dashboards expose, with the exact computation.
Calls dispatched
Count of distinct calls filled by your office in the period. One call = one show request, regardless of how many days or how many crew it required.
Days on call
Sum of member-days dispatched in the period. A 6-person, 3-day install contributes 18 days. Counts each crew-day regardless of hours worked.
Members dispatched
Count of unique members who worked at least one day in the period. The denominator behind member-days-distribution and fairness metrics.
Office turnaround
Time from request received (email parsed, text received, booking page submitted, voicemail transcribed) to offers sent to the first crew member. Measured in minutes. The speed of the office.
Crew confirmation
Time from first offer sent to last slot confirmed. Measured in minutes. The speed of the crew network.
Override rate
Percentage of dispatches that required the BA to bypass the system's ranked recommendation. Every override carries a required reason field — captured for the audit log and used by the TD to learn the local's specific exceptions.
OOW order respected
Percentage of dispatches that went to the system-ranked top eligible member. The inverse of override rate, scoped to OOW-relevant calls.
Days distribution by tier
Average days dispatched per member, segmented by tier (Journey, Apprentice, etc.). The fairness metric — surfaces whether days are spreading evenly within tier per the local's rules.
Start-time distribution
Histogram of call start times by hour of day. Operational planning signal — tells the office what its rhythm actually is.
Day-of-week distribution
Bar chart of calls by day of week. Tells the office whether weekend premiums are firing as expected.
Call shape
Distribution of calls by day count: 1-day, 2–4 day, 5+ day. Tells the office what kind of work it's actually moving.
baseline methodology
Comparing to before Dispatch.
Every claim that Dispatch “saved time” or “reduced threads” requires a baseline. We capture baselines two ways:
Self-reported during onboarding
The BA, dispatcher, or shop owner answers a structured intake during onboarding: How long does it typically take to fill a 6-person stagehand call right now? How many threads (texts, emails, voicemails) does that take? How long does it take to defend a dispatch decision in a grievance hearing today?
Captured. Timestamped. Attributed to the person who answered. Treated as the baseline going forward.
Shadow period (optional)
For 30 days during onboarding, Dispatch runs alongside the office's existing workflow. Office turnaround, threads-per-call, and time-to-fill are measured in both systems. The delta is the evidence.
Shadow period gives harder data. Self-reported gives faster onboarding. Both are captured for design partners. Either is enough for the org's own dashboard.
three views, three jobs
What the dashboards show.
for crew
The member's contribution + fingerprint
Calls accepted, days worked by classification and venue, response rate. TD citations. Contribution patterns. The crew member's own data, scoped to them.
See the crew viewfor the employer
Booking history + learned patterns
Calls placed, days booked, gig type distribution. Per-show recall. The patterns the TD has learned about this employer's shows. The pre-fill on next year's gala.
See the employer viewaggregates
What aggregates, and what doesn’t.
Anonymized aggregate metrics — opted in per org, per the consent audit log — contribute to industry-wide observations:
- Gear rental ranges by category (min, quartile, median, max), floored at n ≥ 5
- Time-to-fill benchmarks by gig type and crew size
- Classification-mix patterns by show type
- Office turnaround distribution across all opted-in orgs
the wall
No individual call, member, vendor, employer, or show is identifiable in any aggregate. No raw event data leaves the contributing org. Aggregates are computed against the consent ledger — if an org opts out, their contributions stop flowing forward immediately.
Why aggregates exist: a producer walking into a negotiation knowing the median and spread of comparable gear packages has better information than a producer guessing. An org seeing where it sits in the office-turnaround distribution can benchmark itself against the industry without exposing its own data to anyone.
honest limits
What the numbers can’t tell you.
They can't measure what didn't happen
The grievance that didn't get filed because the audit log was airtight isn't a number on the dashboard. The bid you didn't lose because the office turnaround was 23 minutes instead of 2 hours doesn't show up either. The biggest wins are often invisible.
They don't replace operator judgment
A 100% OOW-respect rate doesn't mean the BA isn't doing their job; it means the system is computing ranking correctly per the local's rules. A high override rate doesn't mean the BA is wrong; it means the TD hasn't finished learning the local. Both numbers are inputs, not verdicts.
They mature with use
Month one numbers reflect month one usage. The dashboards get more meaningful as the corpus deepens — typically by month six the patterns are stable enough to make strategic decisions from. Read year-one numbers with appropriate context.
Want to see your office’s numbers, before and after?
Request a private walkthrough →or email allen@soxal.co
next up
How those numbers improve over time — every call, every override, every correction, training the TD that runs your office.