Client€3B Global Speciality Chemicals Company
SprintAn 8 week sprint
Fieldwork1,862 deliveries analysed · 14 explanation codes traced · 9 platforms scanned
Supply Chain and Sustainability
On time against the wrong date
The carrier scorecard said 92 per cent on time. Customers were still ringing to say deliveries were late. Both were true, because the two sides were measuring against different dates. Against the date the customer had actually been promised, the same deliveries were on time 70 per cent of the time.
1,862
Deliveries analysed after the data was cleaned
22pp
Gap between the carrier measure and the customer promise
45%
Of pick ups happened early, flattering the delivery figure
4
Measurement changes agreed with the logistics partner
A note on sourcing. Every figure here comes from a SprintlyWorks client engagement. Clients are described, never named. Where a figure is identified, modelled or indicative rather than banked, the line says so.
01
Two measures, one of them the customer never saw
The outbound network is large and fragmented. In the region in scope the client works with 400 to 500 carriers running 100,000 to 120,000 deliveries a year, four fifths of it bulk in tank trucks and the rest packed goods. Those are the client own figures.
The transport management system provided by the logistics partner offered a visibility tool, but drivers were only required to submit a delivery confirmation, and confirmations were arriving days late against a 24 hour requirement. Nothing could be seen until the carrier had put it into the system. So the only thing being measured was the thing the carrier chose to report, after the fact.
The brief had two halves. What visibility options exist and what do they cost, and how should on time delivery be defined so that it means what a customer means by it.
02
The data had to be fixed before it could be read
The delivery report was missing the fields the question needed, including the originally requested delivery date and the actual delivery date. Fixing the report was the first deliverable rather than a preliminary.
The scale of the gap shows in the funnel. 3,769 external deliveries in the region reduced to 1,905 once records with missing data were excluded, and to 1,862 after repeat shipments were removed. Everything on this page rests on that 1,862.
The first promise date, the one the customer is actually given, did not exist in a usable form at all. Two ways of reconstructing it were tested and both returned the wrong date. The fix was structural: lock the first agreed delivery date in the enterprise system at the point it is agreed, and add it to the report.
03
Early pick up was hiding the problem
Carriers were collecting early in 45 per cent of cases. Having collected early, they then delivered on time against the schedule in 92 per cent of those cases, which produces an excellent carrier score.
Measured against the date the customer had been promised, the same early collection cases delivered on time only 75 per cent of the time. And where the pick up happened on schedule rather than early, on time delivery was 93 per cent by the carrier measure and 66 per cent by the customer promise, with a quarter of them late.
Transit time was being underestimated in the plan. Against the partner planned transit time, 46 per cent of deliveries took longer than estimated. That is not an exception rate, it is close to half.
Two scorecards
The same 1,862 deliveries, measured two ways
On time counted strictly, and on time counted with early deliveries included as successes. Baseline measurements of the existing situation, not improvements delivered.
| Measure | On time only | On time plus early |
|---|---|---|
| On time delivery, carrier measure | 92% | 95% |
| On time delivery, customer promise | 70% | 78% |
| On time pick up | 49% | 94% |
| On time shipment | 53% | 74% |
How to read this → The second column is the reason nobody had noticed. Counting early as a success turns a 49 per cent pick up record into 94 per cent.
SprintlyWorks analysis
04
Nobody was explaining the exceptions either
The partner system carries fourteen explanation codes for a delivery that misses its date. In practice they were mostly unused. Of 66 early deliveries, 46 had no explanation recorded at all. Of 87 late deliveries, 28 had none, and only 18, one per cent of the whole sample, were coded as genuinely too late.
Only the too early and too late codes are treated by the partner as the carrier fault, which means the fault rate on that system is one per cent by construction.
A tolerance test made the point another way. Allowing plus or minus one day would add 4.5 percentage points to the delivery measure and 39 percentage points to the pick up measure. Those are modelled restatements of the same sample, not gains, and they show how much of the performance question is definition rather than operation.
05
What actually changed, and what was left open
Four measurement changes were agreed. The first agreed delivery date gets locked in the enterprise system and added to the report. The partner stops counting early deliveries as on time. Carrier explanation codes go into the report so the pattern behind late and early can be seen. And pick up accuracy is accepted as unmeasurable objectively until a scanning or real time solution exists.
The visibility question was deliberately not answered with a vendor. Nine platforms were scanned and twelve capabilities compared across the three closest, and the conclusion was that they offer much the same features and differ mainly in which transport modes they cover. The decision was framed back to the client as a question about appetite and investment, not as a shortlist.
There is no measured improvement in this work and none is claimed. Every figure here is a baseline of the situation as it was. What it produced was a definition of on time that a customer would recognise, and the honest news that the existing one was not it.
Download the full case study
Have a similar requirement?
Contact us today to learn more about on-demand workforce and accelerate development on your most pivotal projects!
Quick Reads for Big Impact
Featured Case Studies
Supply Chain & SustainabilityEight fixes for a fault everyone already knew
3 September 2026Read ›
Supply Chain & SustainabilityThirty indicators for something everyone called a culture
1 September 2026Read ›
Supply Chain & SustainabilitySixteen factories, sixteen ways to fix a machine
29 August 2026Read ›Quick Reads for Big Impact
Accelerating Success for Enterprises in 20+ Geographies
Launch Your Sprint with
Define your project, connect with top-tier consultants, and start making progress fast.