CARRIER CASE STUDY
From Manual Service Reports to a Trusted View of Carrier Performance
CARRIER PERFORMANCE | LOGISTICS
Nine thousand shipments. One reliable picture.
A Canadian carrier moving over 3,000 shipments a month needed to know two things: which of their numbers were trustworthy, and what their service levels actually were. The answer required two separate pieces of work — and a deliberate order.
LOGISTICS
THE STARTING POINT
Data was coming in from multiple sources. Numbers didn’t always match. Service-level reports existed, but reconciling them was a manual process — and there was genuine uncertainty about which figures to rely on.
Before building anything new, the first question had to be answered: which data can we trust?
ENGAGEMENT ONE
Which numbers can we trust?
Three months of shipment data — reviewed record by record.
The work established data provenance: where each record originated, how records from different sources related to each other, and which discrepancies were systematic versus incidental. The output wasn’t a dashboard. It was a clear account of what the data meant and which figures were reliable enough to build on.
This step is not optional. Automating an unreliable process doesn’t fix it — it produces questionable information faster.
ENGAGEMENT TWO
How are we really performing?
With trusted data established, the service-level engine could be built.
The rules were defined operationally: what counts as a business day for this carrier, how to classify a shipment as on-time, late, in-transit, or missing. Those classifications were applied across the full shipment dataset — 9,500+ shipments validated against existing reports, shipment for shipment.
The result: A service-level dashboard the carrier’s team could stand behind.
WHAT CHANGED
9,500+ shipments classified and validated against existing reports
Service-level performance visible at a glance — by lane, carrier partner, and time period
On-time, late, in-transit, and missing statuses defined by operational rules, not interpretation
Business-day calculation standardized and applied consistently across the dataset
Data provenance documented: the carrier knows where each number comes from
Discrepancies between source systems identified and accounted for
Reporting that previously required manual reconciliation now runs automatically
A foundation the team trusts enough to use for operational decisions
THE LESSON
Trustworthy operational data doesn’t come from better tools. It comes from understanding the data you have — where it originates, what it actually measures, where it disagrees with itself — before you build anything on top of it.
The sequence matters: establish what can be trusted, define the operational rules, validate against what you know, then automate. Skipping the first two steps means the dashboard is fast, but the numbers are still questionable.
Running on manual reporting or uncertain data?
If your team is spending time reconciling numbers instead of acting on them, the problem is usually upstream from the dashboard.