The daily Meesho routine that stops breaches.
Most missed dispatch windows are not caused by slow packing. They are caused by processing orders reactively all day instead of clearing them in one early batch. Here is the sequence that fixes it.
- One early batch beats reacting to orders all day — every order inherits maximum runway.
- Sort by time remaining, never by order date. Windows differ between orders.
- Separate blocked orders immediately so one stock-out does not stall twenty parcels.
- The most-missed step is confirming the handover scan actually happened.
- Robnu sequences the batch by urgency and tracks every clock. Free while we figure out pricing.
There is nothing clever about this routine. It is deliberately boring, because the thing that breaks marketplace operations is not difficulty — it is the accumulation of small decisions made repeatedly under time pressure. A fixed sequence removes the decisions.
Batch, do not react
Handling orders as they arrive feels responsive. In practice it means switching contexts dozens of times a day, doing each step badly, and giving your latest orders the least runway. Batching means downloading all labels once, packing in one pass, and handing over as a single consignment.
The runway argument is the decisive one. An order processed at 11am has hours of buffer before its window closes; the same order picked up at 7pm may have almost none. Nothing about your packing speed changes that — only when you start does.
Six steps, same order, every day
Run them in this order. The sequence matters as much as the steps.
Sort by deadline
Open the pending queue and order it by time remaining. Anything already in breaching soon jumps the queue.
Triage blockers
Pull out stock-outs and label failures. They get handled separately, not in the flow.
Labels in one pass
Download and print everything together. One printer session, one stack, no switching between the panel and the packing table.
Pack and weigh
Record the weight as you go. That record is your evidence in any later weight dispute.
Manifest and hand over
Get the manifest signed. This is your proof a handover occurred if a courier delay is later attributed to you.
Close-out check
Confirm every dispatched order has an actual scan, and that nothing new has drifted into the warning zone. Two minutes, saves tomorrow’s penalty.
When the routine outgrows the person
This checklist works, and it keeps working until volume outpaces the amount of state a human can hold. At that point the failures are not carelessness — they are orders that genuinely fell out of attention while attention was elsewhere.
Robnu is an agentic OMS. It holds the queue sorted by urgency, tracks the dispatch clock on every open order, flags blockers before they stall a batch, and tells you when a handover has not been scanned. The routine stops depending on anyone remembering it.
You sell. Robnu runs the rest — and makes sure every rupee is paid correctly.
Order processing, answered
In one batch, early, in a fixed sequence — rather than reacting to orders as they arrive through the day. Batching means you handle each step once for all orders instead of context-switching repeatedly, and starting early gives every order in that batch the maximum possible dispatch runway. Reactive processing is the single biggest cause of avoidable SLA breaches.
As early in your working day as your order flow allows. An order processed mid-morning has substantially more runway before its window closes than the same order processed in the evening. If you can only run one batch, run it early; if you can run two, make the second one a catch-up rather than the main event.
By deadline, always. Dispatch windows differ between orders, so the oldest order is not reliably the most urgent. Sorting by time remaining rather than order date is a small change that prevents a specific and common failure — breaching a newer order while working carefully through older ones.
Confirming the courier handover actually got scanned. Sellers reliably print labels, pack parcels and mark orders ready — then assume pickup happened. An order that is packed but never scanned keeps counting down toward a breach, and it is the most frustrating way to lose an order because all the real work was done.
Separate them out immediately rather than letting them stall the batch. A stock-out and a label that will not download need different responses, and neither should hold up the twenty orders behind them. Deal with the flowing orders first, then work the blocked pile with whatever time remains.
It is strongly worth having. A manifest that the courier signs is your evidence that a handover took place at a given time — which is exactly what you need if an order is later marked as a seller dispatch delay when the courier failed to collect. Without it, a waiver request is your word against a scan record.
The sequence stays the same; what breaks is the human tracking layer. At five orders a day you can hold the state in your head. At twenty-five you cannot, and the failures shift from effort problems to visibility problems — orders forgotten rather than orders skipped. That transition is where a system stops being optional.
Two things: that every order you intended to dispatch has an actual handover scan, and that nothing has quietly appeared in the breaching-soon state while you worked. A two-minute close-out check at the end of the batch catches almost everything that would otherwise become tomorrow's penalty.