Logistics reports — eight screens and what they measure
Choose a report based on the decision you need to make. Check the unit, period, and completeness of the data before comparing results.
Find the reports under Reports → Logistics. Each of the eight screens supports a different decision. For daily work, start with Problem orders and Logistics overview.
Choose a screen
| Report | Who uses it and why | What it shows |
|---|---|---|
| Logistics overview | A logistics manager chooses an area to investigate | Parcel volumes, tracking coverage, times, on-time performance, and examples of problematic shipments |
| Lead time | A warehouse manager compares the wait before dispatch | Time from an order arriving in Noxti to its recorded pickup or dispatch |
| On-time performance by cut-off | A warehouse manager checks the promise “orders received by 2 pm ship today” | Share of orders packed or picked up before the cut-off schedule deadline; late and overdue-open orders |
| Carrier performance | A carrier manager checks services and routes | Measurable transport time, deliveries, failures, and carrier comparisons |
| SLA report | Logistics and customer service check a selected time target | On-time delivered parcels, plus overdue-open parcels and results for resolved or overdue cases |
| Transport anomalies | A shift lead or integration owner verifies problem signals | WMS stage backlogs, tracking, scans, and import quality; causes and suggested actions |
| Problem orders | A warehouse lead chooses orders to unblock | Current backlogs against the shipping deadline or age threshold, including partial shipments |
| Advanced analytics | An analyst compares periods; a planner looks at an indicative workload | Year-over-year and week-over-week comparisons and an experimental forecast of incoming orders |
Set the period and check the data
- Choose a period and available filters: marketplace, carrier, country, or shipment type. The default type is SEND, meaning outbound shipping. Returns are a different process and need separate interpretation.
- In period reports, the range usually means the date the order entered Noxti. When recalculated, the result includes the parcel’s known history, including a delivery after the end of the selected period. The trend groups the same population by the order-entry date.
- The range limit is 186 days, including both the first and last day. Lead time, Carrier Performance, and SLA screens allow a custom range. Non-working options for a year or all history have been removed. Separate rules for current backlogs are described below.
- Read the report’s description of its goal and population. In the four parcel reports, check the counts of parcels, parcels with tracking, valid transport-time measurements, and the source of the transport start. Orders without parcels are shown separately; they do not create artificial parcel measurements.
- Check when the report was calculated. The latest tracking event in the selected group is a separate piece of information; it does not confirm when synchronisation last ran.
The cache lasts 10 minutes, or 5 minutes for anomalies and problem orders. The refresh button recalculates the report; it does not fetch new data from the carrier.
What the times mean
The measurement starts when the order enters Noxti, not automatically when the customer buys it. Time before import is not part of measured E2E.
- Lead time: entry into Noxti → pickup status; if that is missing, the recorded dispatch date can be used. The data-quality information distinguishes these sources.
- Transport time: the same pickup or dispatch start → delivery. A delivered status by itself is not enough to determine transport time.
- E2E: order enters Noxti → delivery.
These are calendar hours. The ≤24 h, ≤48 h, and ≤72 h metrics do not mean the next business day. The unit for these measurements is a parcel; one multi-parcel order may contribute several measurements.
The E2E split between pre-shipment time and transport time is calculated for parcels with both valid intervals. It helps identify an area to investigate, but does not assign blame to the warehouse or carrier. Before dispatch, the order may be waiting for payment, stock, readiness, or pickup. A recorded dispatch is also not necessarily the exact time the carrier physically took possession.
How to read the time target in the SLA report
The selected number of days is an E2E target from import to delivery. It is not automatically the delivery promise communicated to the customer.
On-time delivered is the share of delivered parcels with a valid E2E measurement that arrived within the target. A separate result covers resolved or overdue cases: measurable deliveries, failures that can be assessed, and parcels still open after the target has passed. Open parcels still within the target and cases that cannot be assessed are shown separately.
Example: 10 parcels were delivered on time, while 90 remain open past their deadline. On-time performance among delivered parcels is 100%, but the result for 100 resolved or overdue cases is 10%. Read the denominator for each metric.
A missing valid measurement is shown as no data, not 0 h or 0%.
Problem orders: a list to act on
The list excludes cancelled orders and uses the shipping deadline if one was recorded. If there is no deadline, it uses the selected age threshold from the time the order entered Noxti. Each order shows the deadline source and how long it has been overdue.
Pay particular attention to orders with no parcel, no tracking, no dispatch confirmation, and partial shipments. Completeness is assessed from recorded shipments and parcels; the report does not prove that every required unit of every product was sent. You can optionally show orders dispatched late — an old order that was shipped quickly does not become late just because it is old.
Created in the last defaults to 90 days. All dates finds older backlogs; this is a separate mode from the period-based parcel reports. A minimum waiting-time filter, sorting, and pagination are available. Counts cover the full filtered result, not just the visible page. Order values are separated by currency.
Anomalies: signals to investigate
The five categories cover the warehouse, carriers, tracking, scanning, and data quality. WMS backlogs reflect the current state, regardless of the selected date range. The rules account for queue activity, cancellations, pending operations, and skipped process stages.
Other rules analyse the group of orders from the selected period; scan rules analyse scans from that period. The counter includes every detected case; the list shows up to 50 examples per category and indicates how many were omitted. One case can represent a group of scanning errors, so the number of cases should not be treated as the number of orders.
A suggested action helps start an investigation. The report does not record an owner assignment or case closure. No alert does not confirm that the whole process is correct.
Experimental analytics
The WoW comparison covers two non-overlapping weeks ending on the selected date. YoY compares the selected range with the corresponding range a year earlier. In both comparisons, the newer orders have had less time to be delivered, which can affect the result.
The forecast is the average order inflow over the four weeks ending on the selected date. It covers the next seven days, not the number of future deliveries. It includes the marketplace; selecting a carrier or non-standard shipment type narrows it to orders with the corresponding shipment assigned. The default SEND also includes incoming work without an assigned shipment.
Forecast accuracy has not been tested on earlier periods. With no history, there is no forecast. First-attempt delivery success (FDA) is unavailable because the source does not have agreed attempt events and outcomes. The absence of a failed-attempt phrase in tracking does not confirm that delivery succeeded on the first try.
More interpretation rules: Logistics report pitfalls.
Want to see this with your orders? We’ll show you NOXTI with your sales channels and warehouse.
Book a demo