Reading a 3PL Scorecard Before You Sign the Rate Increase
Your 3PL's monthly scorecard shows OTIF at 89% (96% last quarter), order fill rate 97.5% (flat), inventory record accuracy 99.2% (flat), average order cycle time inside the DC at 4.2 hours (3.1 hours last quarter), and carrier pickup on-time at 94% (97% last quarter). Customer complaints are concentrated in one region. Order volume is up 18% year over year; headcount is unchanged. The 3PL account manager proposes adding 12 pickers at $3,400 per month each, citing the cycle-time increase, and asks for a 4% rate increase to cover it. Your contract sets a 95% OTIF target with a 2% rebate at risk.
1. What does the evidence indicate about the constraint, and what does it not tell you?
Read the worked answer
What it tells you: fill rate is flat (97.5%) while OTIF fell seven points. Orders are being shipped complete but late - that is a time failure, not an availability failure. So inventory investment is not the answer. Cycle time inside the DC rose 35% while volume rose 18% with flat headcount - a plausible capacity constraint in picking or packing.
What it does not tell you:
- Whether the lateness occurs at DC release or in transit. If late orders cluster by region, it is a carrier/lane problem, not a labor problem.
- Whether the carrier pickup miss (94%) is caused by the DC not having freight staged - in which case the carrier metric is a symptom of the same warehouse constraint, or by carrier tender rejection.
- Whether the cycle-time rise is pick labor, dock-to-stock delay, or appointment dwell.
A 4.2-hour cycle time is not by itself a service failure. Before approving 12 heads ($40,800/month, $489,600/yr - far more than the 2% rebate), decompose the miss.
2. List the specific data cuts you would demand before responding to the proposal.
Read the worked answer
- OTIF split by component: order fill %, ship-on-time (release vs. committed ship date), and transit-on-time, each by region and lane.
- Late orders by ship date: does the late tail sit on a few peak days, or is it uniform? Peak-day clustering points to labor; uniform lateness points to process or systems.
- DC release time vs. carrier appointment time for each late order - who was late to the dock.
- Carrier tender acceptance and pickup performance by lane, with the reason codes for missed pickups.
- Lines and units per labor hour by shift, plus overtime and temp-labor usage, to test whether the constraint is hours or productivity.
- Order profile mix: average lines per order and small-order share, since rising small-order mix explodes pick time even at flat unit volume.
- Backorder age and past-due backlog to confirm fill rate is genuinely stable rather than being propped up by substitutions.
Ask for 90 days of raw order-level data, not the summary scorecard.
3. How would you restructure the metric set and the rebate clause going forward?
Read the worked answer
- Break OTIF into auditable components with separate targets: order fill >= 98%, ship-on-time >= 97%, transit-on-time >= 96%. A single composite number lets each party blame the other.
- Measure at order-line level and weight by customer value, so a missed line on a top account is not diluted by thousands of clean small orders.
- Define exclusions tightly: customer-caused, carrier-act-of-God, and forecast-error-driven events must be codified, or every miss becomes a dispute.
- Require root cause and corrective action on every miss, with an aging log.
- Use a rolling three-month rebate rather than a single month, so one peak week neither triggers nor hides a penalty.
- Make capacity a joint obligation: a volume forecast shared 8 weeks out, with a defined surge rate agreed in advance. Then the 12-picker question becomes a funded, forecasted capacity decision rather than a surprise rate increase after a service failure.