Independent IFS Cloud practice · Supply Chain
The on-time delivery metric that lies and the outliers it is hiding
Key takeaways
- An aggregate on-time delivery metric can read 94 percent and still hide the three orders that mattered most that month.
- Averages weight every order equally — a late case of packaging counts the same as a late shipment that stops a production line.
- The outliers are cheap to find once you weight by consequence — value, criticality, customer tier — instead of just counting lines.
- A Custom Event that flags late critical orders separately turns the KPI from an answer into a question worth asking every month.
On-time delivery: 94 percent. The dashboard is green. The monthly review notes the improvement and moves on. What the number does not show: the six percent that were late included the three largest orders of the month. Two of them stopped a production line. One cost a customer. None of that is visible in a single aggregate figure, because 94 percent is an average, and averages are built to erase exactly the kind of detail that would have made someone act sooner. Aggregate metrics are not wrong. They are honest about the wrong thing — they tell you the average, and in supply chain the average is rarely what does the damage. This article covers why on-time delivery in particular hides its worst cases, what that costs, how to surface the outliers inside IFS Cloud, and how to build a rule that treats a late critical order differently from a late routine one, without replacing the KPI you already report.
1.Why does the average hide the orders that matter?
On-time delivery is calculated the same way almost everywhere: on-time lines divided by total lines, expressed as a percentage. Every line counts exactly once, regardless of what it was carrying. A pallet of low-value consumables that arrived a day late counts the same as the single component that stopped a production line for a week. The formula has no concept of consequence, only of timing.
That is not a flaw in the formula — it is a summary statistic doing exactly what summary statistics do. The problem starts when the number is treated as the finding rather than the starting point. A team that watches the percentage month over month can genuinely improve the average while the handful of orders that actually cost money stay exactly as late as they always were, simply outnumbered by everything that shipped on time.
2.What does the hidden six percent actually cost?
The gap between the metric and the damage is where the real cost hides. It shows up downstream, in places the on-time percentage never looks:
- Production stoppages. A single late critical component can idle a line for days, at a cost the monthly average will never register.
- Customer loss. One late order to the wrong customer at the wrong moment does more damage to the relationship than fifty on-time deliveries repair.
- Hidden pattern. The same handful of suppliers or part categories often account for the worst outliers month after month, invisible while buried in the average.
- False confidence. A green dashboard tells the business everything is fine, which is precisely when the next large order is allowed to slip unnoticed.
Ninety-four percent on-time is not a result. It is an invitation to look at the six percent and ask which of them actually hurt.
3.How do you find the outliers in IFS Cloud?
IFS Cloud already carries what is needed to separate the outliers from the noise: order value, customer, part criticality, and promised versus actual delivery date all live on the same records the on-time percentage is calculated from. The aggregate simply never asks those fields a question.
Filter late lines by order value above a threshold, by a criticality flag, or by top-tier customer, and the six percent stops being an undifferentiated tail — it becomes a short, specific list of the deliveries that were actually expensive to miss. That list is usually a fraction of the size of the full late report, and far more useful.
4.Aggregate KPI versus outlier-weighted monitoring
The fix does not replace the on-time percentage — leadership will keep asking for it. It adds a second, narrower view built on the same automation pattern behind other SCM alerts in IFS Cloud, watching for the late orders that were never supposed to be treated the same as the rest.
| Aggregate KPI | Outlier-weighted monitoring | |
|---|---|---|
| What it measures | Percentage of lines delivered on time | Which late orders actually mattered |
| Weighting | Every line counts the same | Weighted by value, criticality, customer tier |
| What it hides | The orders that stopped production | Nothing — each is flagged on its own |
| Reviewed | Once a month, as an average | The moment a threshold is crossed |
| Fails when | The average stays green while a handful of misses cost the most | Rarely — each miss is visible individually |
None of this argues for abandoning the percentage. It argues for not stopping there.
5.How do you roll it out safely?
Building the second view is a modest project, not a re-platforming exercise.
- Agree what counts as critical. Set the value threshold, the customer tier, or the part flag that defines an order worth watching separately.
- Run in dry-run. Log which late orders would have been flagged over the past few months and check the list against what actually went wrong.
- Tune the threshold. Widen or narrow the criteria until the flagged list matches the orders that genuinely deserved attention, not everything or nothing.
- Pilot on one category. Apply the rule to a single product line or customer segment first, and confirm it catches what the team already knew mattered.
- Review alongside the monthly KPI. Bring both numbers to the same review — the average for the trend, the flagged list for the outliers that need a decision.
Built inside the IFS Extensibility Framework with standard Custom Events and Workflows, the flagging rule sits alongside the core reporting without touching it — update-safe, with nothing here for the next release to break.
6.Frequently asked questions
Why can 94 percent on-time delivery still be a problem?
Because the percentage treats every late order as equally bad. A late shipment of low-value stock and a late shipment that stops production both count as one miss out of many, so the average can look healthy while the handful of misses that actually cost money go unnoticed.
How do you decide which orders are “critical” enough to weight?
Most teams start with what they already know: a value threshold, a small list of top-tier customers, or a criticality flag already used in planning. The dry-run period usually confirms or corrects the first guess quickly.
Does this replace the on-time delivery KPI we already report?
No. It sits alongside it. Leadership keeps the trend line; the operational team gets a short list of the misses that actually deserve a conversation.
Is this update-safe?
Yes. It uses standard Custom Events, Workflows and PL/SQL inside the IFS Extensibility Framework, with no modification to core on-time delivery reporting.
7.About the author
Dariusz Myśliwiec — 25+ years in ERP and supply chain, 17+ on IFS (Apps 7.5–10 and IFS Cloud). IFS Certified Associate Consultant. PRINCE2® 7. Independent practice based in Kraków, delivered remotely across Europe and globally — you talk to the consultant who builds it, not a sales layer.
Selected clients: Fugro · LGC · BVI Medical · Betafence (PRÆSIDIAD) · Barlinek · NGK Ceramics · Newag · Oleofarm.
IFS is a registered trademark of IFS AB; this practice is not affiliated with IFS AB.
Look past the average
If you suspect your green on-time delivery number is hiding a handful of expensive misses, a dry-run week will show you exactly which orders they are. On a 30-minute call I’ll map the outlier rule to your critical orders — fixed price, dry-run before anything sends.
