Why the Site Manager's Report and the Client's Report Show Different Numbers
In the same month at the same warehouse, the site manager's report can show 99.8% shipping accuracy while the client report shows 98.8%. Neither report is wrong. Take 1,000 orders (4,000 lines, 20,000 units) with 12 orders in error, covering 15 lines and 40 units. Accuracy is 98.8% by order, 99.6% by line and 99.8% by unit, and all three figures are correct.
The real problem is definition. Start dock-to-stock at truck arrival and a receipt takes 6.5 hours. Start it when unloading finishes and the same receipt takes 4 hours. Add manual spreadsheet consolidation and the monthly report shows up in the second week of the following month. If one formula breaks along the way, nobody catches it.
Core KPIs and How to Define Them Precisely
Inbound, Storage and Outbound Metrics
Three Things Every Definition Must State
Each metric gets one line each for its numerator, denominator, and the events that start and stop the clock. For mis-shipments, decide whether you count only errors confirmed by client claims or also errors caught in internal QC. For on-time shipping, decide whether orders delayed by the client come out of the denominator. Skip this agreement and people drift back to spreadsheets even after the dashboard is live.
Building KPIs from WMS Event Data
A KPI is the gap between two timestamps. At minimum, record eight events per receipt or order: arrival, unloading, inspection complete, putaway complete, release, picking complete, packing complete and loading complete.
Finding the Gaps in Event Capture
Arrival and unloading often happen outside the WMS, at a gate system or in a paper log. Measure the capture rate of each event first and the gaps show up. For example, picking completion may be recorded 100% of the time while loading completion is recorded only 60% of the time. Cycle time calculated from that data covers only the orders whose loading was scanned, so the result is skewed.
Flag Corrected Data Separately
When a missing timestamp is entered after the fact or filled in from the next event, set a correction flag. Metrics that rely heavily on corrected data should show a confidence note in management reports.
Dashboard Design Principles
Separate Operations Screens from Management Reports
The floor screen shows only what changes what people do in the next 15 minutes, such as waves at risk of missing cutoff or pallets waiting for putaway. Daily and weekly management views cover performance against target and trends. Put both on one screen and nobody uses it.
Targets, Trends and Drill-Down
Give every metric a target line, a trend for the last 8–12 weeks, and a drill-down to the list of exceptions. When on-time shipping drops, one click should show which orders were late and at which step, picking or loading. Comparisons by client, zone or shift mean something only when every value uses the same definition.
Turning KPIs into Real Floor Improvement
Implementation Sequence
If you build the dashboard first, the arguments about the numbers just move onto the screen.
POLYGLOTSOFT helps design logistics data systems from start to finish: WMS event-data assessment, KPI definition documents, new scan points, and operations and management dashboards. If every report in your warehouse tells a different story, we would be glad to talk.