Efficiency Thinking Model

Diagnose the Flow, Not the Headcount

Adding labor to a broken process only amplifies the waste. The first step is to map the actual picking path — not the one on paper. Walk behind a picker for two hours and record every step, every reach, every pause. You will likely find three systemic leaks: motion waste (walking for supplies), waiting waste (queueing at the packing station), and search waste (hunting for items due to poor slotting).

Apply Munger’s Inversion

Ask: What would guarantee we miss the cutoff every day? The answer: misaligned inventory layout, no batch picking, and pickers acting as their own material handlers. So, invert it — create a dedicated “supply runner” role for one shift, even if just for a trial week. This removes the most visible waste and lets you measure the true picking rate.

Measure the Constraint

Track pick rate per hour before and after the runner is introduced. If output still doesn’t move, the constraint is downstream — likely packing or label printing. Do not add more people until you have isolated the bottleneck. One targeted fix beats two hires every time.

Mapping the Patient Journey — Start With the Queue, Not the Staff

Begin by documenting every step from arrival to checkout, with timestamps. Walk the floor yourself for one full day; do not rely on staff recollections. You will likely find the bottleneck is not the front desk’s effort but the sequence of tasks.

The Hidden Structural Delays

• Check-in and insurance verification are serial, not parallel. Staff pause check-in to answer a phone call, then resume verification. This creates a queue of half-finished tasks. Separate the two: one person owns the phone, another owns the desk.
• Open slots are not truly open. They exist on the schedule but are blocked by pre-visit tasks—insurance eligibility checks, intake forms, or room prep—that were never scheduled as work. Treat each slot as a system with prerequisites, not just a time block.

Apply Munger’s Inversion

Ask: What would guarantee a 40-minute wait? Answer: multiple staff handling every task, no triage of patient urgency, and no buffer for phone interruptions. Invert that—assign single ownership per task, batch phone calls into 15-minute windows, and verify insurance the day before.

Next Step

Map the journey on paper, mark every handoff and wait, then eliminate one handoff per day. Measure wait time weekly. The system, not the people, is your lever.

Trace the Value Stream, Not the Symptoms

Your instinct is correct: the bottleneck is rarely the agents. Apply Munger’s inversion — ask what, if removed, would eliminate 80% of ticket volume. Start by mapping the full journey: signup → first login → onboarding → support request → resolution. Tag every ticket by root cause, not symptom. If 60%+ are “setup steps,” your docs and routing are the constraint, not headcount.

Two Immediate Diagnostics

1. Ticket taxonomy audit — Pull 100 random tickets from the last month and classify them: configuration, navigation, billing, or bug. If setup dominates, the onboarding flow is broken.
2. Routing logic review — Are tickets auto-assigned by topic or by agent availability? Topic-based routing surfaces recurring patterns faster and lets you build a knowledge base from real gaps.

Fix the Loop, Then Scale

Before hiring, invest in a self-serve layer: a short video walkthrough, a checklist embedded in the app, and a searchable FAQ built from your top 10 ticket types. Measure deflection rate weekly. Only when self-serve fails should a human intervene. That is the efficiency model: eliminate repetition at the source, not downstream.