Order MS-58102 was packed by 3 PM, sitting in its flyer bag, label on, ready. The manifest got closed at 6:11 PM — eleven minutes after the cutoff — because the seller was at the bank and the helper didn't know the button mattered. The penalty appeared two settlement cycles later as a quiet line item. Nobody was lazy. Nobody was even slow. The parcel was ready. The timestamp didn't care.
A marketplace SLA penalty is unlike every other deduction a seller eats, because it is the only one where evidence is irrelevant. Quality disputes have photos. Payment shortfalls have reconciliation. An SLA breach has a timestamp, and the timestamp is the verdict. That makes this category uniquely unforgiving — and uniquely preventable, because a deadline you can see coming is a deadline you can beat. This guide covers how the clocks actually work on AJIO and Meesho, why breaches happen to diligent sellers, and the prevention stack that makes penalties a non-event.
How the SLA clock actually works
Marketplaces sell buyers a delivery promise, and they subcontract that promise to you as a chain of deadlines. The exact numbers vary by marketplace, category, and your account tier — always read the current policy on your seller panel — but the structure is constant, and it nests:
The acceptance clock starts when the order lands — including the ones that land at 9:14 PM on a Sunday. The dispatch clock is the big one: your order must be processed, packed, and manifested by a cutoff, typically next-day or day-after for fashion. The handover clock closes the loop: the courier's pickup scan needs to match what your manifest claims. Miss any of the three and the system writes the breach automatically — no human reviews it, no one calls you, and the deduction surfaces in a settlement statement days or weeks later.

The detail that catches everyone: the printed cutoff is not the real cutoff. If the manifest closes at 6 PM and your courier pickup historically arrives between 4 and 5, your real deadline is 4. Add a printer jam, a panel timeout, or a helper who's never closed a manifest alone, and the last genuinely safe hour drifts earlier still. SLA discipline is the practice of working backwards from the worst realistic afternoon, not the printed number. Write your own backward chain once — cutoff, minus pickup window, minus packing time for a full day's batch, minus one failure's recovery time — and put that hour on the wall, not the marketplace's.
Why breaches happen to diligent sellers
The comforting story is that penalties happen to careless sellers. The data from sitting with sellers at 5–25 orders/day says otherwise — the causes rank roughly like this (illustrative pattern, but stable across stores we've seen):

Read the top of that chart again: the single biggest cause is orders nobody saw in time — the Saturday-evening order that aged 14 hours before the Sunday panel check. Second is the order MS-58102 story: work done, button not pressed. Third is the courier simply not showing up. Five of the six causes are visibility failures. Penalties are what it costs to run a deadline business on a check-the-panel-twice-a-day information diet.
The courier-pickup cause deserves a special note, because it feels the most unjust: you did everything right and the van didn't come. The marketplace's systems usually cannot distinguish “seller wasn't ready” from “courier didn't arrive” — both look like a missing handover scan. What tilts the occasional dispute your way is contemporaneous documentation: the closed manifest timestamp, a same-day note of the failed pickup, the courier supervisor's acknowledgment. Created that day, it is evidence; reconstructed three weeks later when the penalty lands, it is a story.
Finding the penalties you are already paying
Before fixing future breaches, find the current ones — most sellers underestimate their own penalty rate because the lines are easy to miss. SLA deductions rarely announce themselves; they appear in settlement statements under headings like “breach fee,” “SLA adjustment,” or a bare deduction code, often netted into an order's line rather than listed separately. Pull your last three settlement cycles and search specifically for them. Three cycles matter because penalties cluster: a single bad Thursday — a failed pickup, a helper's day off — can put six breaches into one statement and zero into the next two, and a one-cycle check will tell you a comforting lie.
While you're in there, separate the penalty amount from the penalty pattern. ₹400 of deductions spread across four random orders is noise. ₹400 of deductions that all share a cause — all weekend orders, all one courier, all one helper's shift — is a signal pointing at a single fix. The statement is a diagnostic; read it for shapes, not just totals.
The prevention stack
You do not fix SLA breaches with motivation. You fix them with a stack — four layers, each catching what the previous one missed:
- Layer 1 — process early, on a schedule. If orders sync and process at 6:30 AM automatically, the day starts at zero backlog and the evening orders are already in the morning batch. Most breaches die right here.
- Layer 2 — give every order its own clock. Not “orders are due today” but “MS-58102 breaches in 3h 40m.” A list sorted by SLA headroom, worst first, is the single most useful screen in marketplace ops.
- Layer 3 — alert with margin, escalate without mercy. A notification two hours before breach is actionable; a report after breach is a tombstone. The second alert should be louder than the first.
- Layer 4 — keep the handover honest. Match manifest closure against the courier's pickup scan daily. When the pickup fails, document it the same day — it is the one SLA conversation where you occasionally have a case.

Sale days: when SLA risk multiplies
Everything above gets harder by an order of magnitude during sale events, and it's worth naming why. A sale compresses hundreds of orders into a day or two — which means hundreds of dispatch deadlines land in the same 48-hour band right after. The seller who manifests 25 orders comfortably at 4 PM discovers that 250 orders hit the same cutoff as a wall. Worse, every weak point in the stack fails simultaneously: the unseen-order problem scales with volume, the courier pickup that flexes for 30 parcels doesn't flex for 300, and the printer that jams once a month statistically jams on the worst day.
The sale-day adaptations: dispatch in deadline order rather than order-number order, pack what breaches first; confirm extra pickup capacity with the courier in writing before the event, not during it; and generate all documents in the morning run so the afternoon is purely physical work. Sellers who run sale events on scheduled processing report the same thing — the volume was the marketplace's problem; only the packing was theirs.
The weekly SLA ritual (manual version)
If you run this by hand for now, the ritual is fifteen minutes weekly: pull last week's settlements and search the penalty lines; for each one, find the cause in the chart above — unseen order, late manifest, failed pickup; then fix the category, not the incident. One failed pickup is weather; three failed pickups on Thursdays is a courier conversation. Two late manifests by a helper is a checklist taped above the packing table. The settlement tells you where your stack is leaking — read it as a diagnostic, not a bill.
Where Robnu fits
Robnu is the agentic OMS for AJIO and Meesho sellers, and the prevention stack above is not a metaphor for it — it is the literal architecture. Scheduled processing keeps the backlog at zero, the operations dashboard sorts every open order by SLA headroom, and escalating alerts fire while the parcel can still make the cutoff. The result, in stores we've watched move onto it, is that the penalty line in settlements goes quiet — not because anyone became more disciplined, but because the deadlines became visible early enough that beating them stopped requiring discipline. (Robnu is independent software built for marketplace sellers, not affiliated with AJIO or Meesho.)
Robnu is free for everyone right now — every feature, every order, no card. When paid pricing eventually launches, sellers under 25 orders/day stay free forever, and early users get grandfathered locked rates. The timestamp will always be the verdict. The trick is to stop generating timestamps that convict you.
