What courier dashboards do not tell you
Aggregators give you NDR tools, RTO scores and tracking. What they do not give you is the cost truth. The gap between reports and reconciliation.
If you ship through an aggregator, you already have tools: tracking, NDR management, delivery analytics, maybe an RTO predictor at checkout. A reasonable assumption follows: the RTO problem is instrumented, so it is managed. The dashboard tells a reassuring story. Here is what it leaves out.
The courier and aggregator stack is genuinely good at shipment events: where the parcel is, which attempt failed, what reason code was logged. NDR tooling has improved fast; automated WhatsApp follow-ups and reattempt workflows are now standard features, not premium ones.
The cost truth. The dashboard counts events, not money. Nowhere does it tell you what last month's failed orders cost across both freight legs, packaging, fees, damage, blocked stock and wasted acquisition spend. The RTO rate is shown; the RTO cost is yours to compute from your own numbers, not the courier's.
The reconciliation gap. What the courier charged versus what the rate card says; what was collected versus what was remitted. The aggregator holds one side of that data and you hold the other, and the matching is your problem.
The reason behind the reason codes. The dashboard records "customer refused". It does not record that the customer was never called, or that the address never had a landmark, or that the same customer refused for the fourth time. Root causes live outside the tool.
Whether your fixes worked. You can turn on confirmation calls, NDR automation and risk scoring, and the dashboard will happily show you activity. Did RTO fall for the cohort that got the intervention, against the cohort that did not? That comparison is an analysis, and dashboards do not run it.
This is not a conspiracy; it is incentives. Aggregators earn on shipments, including returned ones. Their tools are built to move parcels efficiently, and they do that well. Answering "what is failure costing you, and which fix pays" is an analytics job on your P&L, and it sits on your side of the table.
The close is unglamorous: export the data, cost it honestly, and follow through. Take 90 days of orders, tag every failed COD order, cost each one across the full stack, group the failures by cause, and price the top fix. Then measure for 30 days after the fix. That loop (reconcile, fix, measure) is the whole method. Tools help at each step, but none of them runs the loop for you.