A stationery brand in Izmir takes an order on Friday at 15:10. The warehouse is two people and they are good at their jobs. The parcel is picked and packed by 16:30, the label is printed at 17:40, and the customer gets a cheerful email: your order has shipped.
Nothing has shipped, and the order's real dispatch time is only just beginning. The courier came at 14:00 and left. The branch stops accepting drop-offs before the last transfer vehicle leaves, and nobody was going to make that. Saturday the pickup request had to be in by mid-morning and no one was in the office to make it. Sunday nothing moves at all. The parcel is collected on Monday afternoon, and the first movement record appears in the carrier's system on Monday evening.
The carrier then does its job well: the parcel is delivered on Wednesday morning. On the carrier's scorecard this is a shipment that took a little under two days, comfortably inside the national average. On the store's dashboard it is an order that shipped on Friday. In the customer's memory it is five days.
Roughly seventy-two of those hours happened before the carrier touched the parcel. That is close to two-thirds of the total, produced entirely inside a business that believes its shipping problem is a carrier problem.
This guide is about order dispatch time — the hours you own. It is a companion to two posts that cover the halves either side of it: estimated delivery dates is about the promise you make at checkout, and carrier performance metrics is about the leg after hand-off. If you are deciding whether to offer a fast delivery option at all, start with same-day and next-day delivery instead.
The Delivery Clock Has Three Parts, and You Only Measure One
The customer runs one clock. It starts when the payment goes through and stops at the doorbell. Everything in between is one undifferentiated wait, and they do not care which company owns which segment of it.
Your business runs the same period as three separate things.
Part one: order to label. Picking, packing, deciding which carrier, generating the label. This is entirely yours. Most stores have a rough feel for it and no number.
Part two: label to first carrier record. The parcel exists, the label is on it, and it is standing on a table. This is also entirely yours, and almost nobody measures it — because the store platform stops caring at "shipped" and the carrier has not started caring yet.
Part three: first record to delivery. The carrier's leg. This is the only part with a number attached, because the carrier publishes it and every dashboard reports it.
The asymmetry matters more than it sounds. Yurtiçi Kargo's own description of its address pickup service says two things worth reading carefully: courier requests are queued in the order they are submitted, and the company's time commitments begin after the parcel is collected. That is not a loophole, it is an honest statement of scope. The carrier is promising you a transit time, not a total time. Where their clock starts is where yours was supposed to stop.
The baseline you are being compared against. The most recent Ministry of Trade figures put the average e-commerce delivery time in Turkey at 42.2 hours, with 13.8% of parcels arriving in under 24 hours, 53.1% in 24 to 48 hours and 26.3% in 48 to 72. Read that distribution as your customer reads it, because they do not know it excludes your dispatch time. A store with a 30-hour dispatch lag and a good carrier delivers in about three days and sits in a part of that distribution its carrier scorecard never shows.
"Marked as Shipped" Is Not "Shipped"
The heading above is the single most useful sentence in this guide. In almost every store, the shipped notification fires when a human clicks something or when a label is generated — not when a parcel physically moves. Those two events can be minutes apart or they can be three days apart, and nothing in the system distinguishes between the two cases.
Where the gap comes from
There are only a handful of causes, and every store has at least two of them.
Labels get printed in a batch at the end of the day, after the courier has already been and gone. Someone drops parcels at the branch after the last transfer vehicle has left, so they sit until the next morning. A label is bought today for a parcel that will not be packed until tomorrow, because the rate was good or the batch was convenient. An order arrives late on Friday and there is no weekend hand-off. Or the parcel is simply waiting on something else — a stock check, an invoice, a customization, a phone call to the customer about their address.
Fix it: for one week, write down the reason every parcel that missed same-day hand-off missed it. One line each. You will find two causes account for most of it, and neither will be the one you assumed.
Why customers feel this more than a delay later on
A delay in transit is invisible until the promised date passes. A dispatch gap is visible immediately, because you announced it.
The shipped notification starts a clock in the customer's head. They open the tracking link that evening and see nothing. They open it again the next morning and see nothing. By the second day they are not asking whether the parcel is late — they are asking whether it exists, and whether you actually sent it. That is a trust question, and trust questions do not arrive as polite emails.
This is also the most expensive kind of contact you can generate. Where-is-my-order questions run at 35% to 60% of all support tickets in most e-commerce operations, and the ones produced by a silent tracking link are the hardest to close, because the honest answer — "it hasn't moved yet" — is the one answer nobody wants to give. Gartner's research on proactive service is blunt about the mechanism: a proactive message that raises a question it does not answer generates contacts rather than preventing them. A shipped email pointing at an empty tracking page is exactly that message.
Fix it: either move the notification to the moment of hand-off, or change what it says. "Your order is packed and leaves our warehouse tomorrow morning" is a message that answers the next question. "Your order has shipped," followed by two days of nothing, is a message that asks one.
How to measure it
You already have both timestamps. The label carries a creation time. The carrier's tracking history carries a first movement record — usually the acceptance or first-scan event. The gap between them is the number nobody in your business has ever looked at.
Three rules for reading it. Use the median, not the average: one parcel that sat for a week will drag an average into meaninglessness while hiding a chronic two-hour problem. Split it by weekday, because Friday behaves nothing like Tuesday. Split it by carrier, because pickup discipline varies enormously between carriers and even between branches of the same carrier.
Fix it: pull thirty consecutive shipments, note both timestamps in a spreadsheet, and compute the median. This takes about twenty minutes and it is the only measurement in this guide you genuinely cannot skip. Everything else is a response to what that number says.
Your Real Cutoff Is the Earliest of Three Clocks
Most stores publish a cutoff time — order by 16:00 for same-day dispatch — and derive it from exactly one thing: how late the team is willing to work. That is one of three constraints, and usually not the binding one.
Clock one: how many orders you can actually pick, pack and label
This is the constraint everyone thinks about, and it is real. It also moves with volume, which is why a cutoff that held all summer stops holding in November. If your team can process eighty orders in an afternoon and a peak-season Tuesday brings two hundred, the cutoff did not change but the truth behind it did.
Clock two: the carrier's deadline for requesting a same-day pickup
This one is published, and almost nobody in e-commerce has read it. It is also usually earlier than the store's own cutoff. As published at the time of writing:
| Carrier | Weekday pickup request | Saturday |
|---|---|---|
| Yurtiçi Kargo | Call center or app during working hours; requests are queued by submission time | Request by 10:30 via call center, or notify the branch by 11:00; Saturday service runs 13:00–17:00 |
| Aras Kargo | Notify the branch by 16:30, or the call center by 15:30 | Call center by 12:00; branches typically 08:30–13:00 |
| DHL eCommerce | System entry by mid-afternoon for same-day collection from e-commerce warehouses | Half-day operation, limited coverage |
Treat that table as a prompt, not as fact: these deadlines change, they differ by branch and by the terms of your own agreement, and the only version that matters is the one your account representative confirms for your address. Sundays, branches are closed.
Fix it: call your rep and ask two questions — what time must a pickup request be in for same-day collection, and what time does the courier actually arrive at our address. The gap between those two answers is a planning constraint you are currently discovering by accident.
Clock three: the last transfer vehicle out of the branch
This is the clock nobody publishes, and for stores that drop parcels off rather than book collections it is the binding one. A parcel handed over after the transfer vehicle has left is a parcel that spends the night in the branch and gets its first movement record the next day. From the customer's side that is indistinguishable from you not having shipped at all.
Fix it: ask the branch what time their outbound vehicle leaves, then set your drop-off routine at least an hour earlier. If you cannot get a straight answer, work it out from your own data — the first-scan timestamps on your last thirty drop-offs will show you the boundary.
Your real cutoff is the earliest of these three, minus a buffer. Publish that one. A cutoff derived from your packing capacity alone is a promise your carrier never agreed to keep.
The Weekend Is Half Your Dispatch Problem
Run the Friday arithmetic once and you will never look at your weekly dispatch numbers the same way.
An order placed Friday afternoon, after the pickup deadline, is collected Monday. That is roughly seventy hours before the carrier's clock starts, on a parcel that was packed within an hour. Add a normal domestic transit and the customer waits close to five days for something that took your warehouse sixty minutes. Meanwhile Monday's orders — placed three days later — go out on a similar timeline, and one of those two customers is going to notice.
Saturday is a half day almost everywhere: pickup requests need to be in by mid-morning, branch hours are short, and coverage is thinner. Sunday is nothing. So a third of the week is either half-speed or stopped, and it sits directly after the day most stores generate a big share of their weekly orders.
There are only three honest responses, and you have to pick one.
Bring Friday's cutoff forward. If the pickup request has to be in by 14:00, then your Friday cutoff is not 16:00 — it is 13:00, and saying so costs you far less than the silence does.
Buy a weekend hand-off. A Saturday morning drop-off, or a Saturday pickup request placed on Friday, moves a whole weekend of orders into Monday's distribution instead of Tuesday's. For a lot of stores this is the single highest-return change available.
Say so on the site. "Orders placed after 13:00 on Friday are dispatched Monday" is not a weakness. It is the difference between a customer who planned around it and a customer who feels misled.
Fix it: split your dispatch-lag median by day of the week. If Friday's is more than double Tuesday's — and it usually is — you have a weekend policy problem, not a warehouse problem.
What Actually Eats the Hours Before the Label
When stores try to speed up dispatch, they generally reach for the warehouse: better shelving, closer pick paths, a second packing bench. Those help, but at 50 to 5,000 orders a month the hours are rarely lost in picking. They are lost in the handful of small administrative steps that sit around it.
Orders that arrive late in your own system
You cannot pack an order you have not seen. If orders reach your shipping process through a manual export, a scheduled sync, or somebody remembering to check a second sales channel, then part of your dispatch time is just latency in your own plumbing. Stores selling on more than one channel lose the most here, and they lose it invisibly — see multichannel order management for the wider version of this problem.
Fix it: measure the time between the order being placed on the channel and appearing in whatever system you pack from. If it is more than a few minutes, that is free time you are giving away.
Choosing a carrier one order at a time
For a store with more than one carrier agreement, every order carries a small decision: which carrier, at what rate, for this destination and this parcel. Done by hand, that decision is thirty seconds of looking things up and a bit of hesitation. At 150 orders a day it is over an hour of pure deliberation, and it happens at exactly the point in the day when the pickup deadline is closing in.
Fix it: turn the decision into a rule before it becomes a daily task. Destination, weight band and service level cover most of it — multi-carrier strategy is where the rate logic belongs, but the operational point here is simply that a rule executes in zero seconds and a judgment call does not.
Address problems discovered at the label step
An address that fails at the label step stops that parcel dead. Someone has to call the customer, wait for a reply, and re-enter the order — and that parcel is now tomorrow's problem, not today's. Around two percent of parcels have an address problem of some kind, and the cost of finding out late is not the correction, it is the day.
Fix it: validate at the point of order, not at the point of print. Failed deliveries covers the downstream damage; the dispatch-time point is narrower — every address you fix before packing is a parcel that does not miss the courier.
Paperwork that blocks the parcel
In Turkey the invoice frequently travels with the parcel, and in more operations than anyone admits, the invoice is the thing everyone is waiting for. If a person has to open an accounting program and re-type order details before a box can be taped shut, your dispatch time includes a data-entry job. Automated e-fatura and e-arşiv issuing removes it entirely, and the time saving usually surprises people more than the accuracy improvement.
Fix it: time the invoice step for one day. If it is more than a few seconds per order, it is not an accounting task any more, it is a dispatch constraint.
The one-by-one label habit
Printing labels individually is the most visible waste and the easiest to fix. The shipping automation post covers the mechanics; the number that matters here is that a manual per-order flow runs to roughly three minutes an order, and a batched one runs closer to twenty seconds. At ninety orders a day that is four and a half hours against half an hour — and crucially, half an hour fits before the pickup deadline while four and a half hours does not.
Fix it: batch by carrier and print once. Then move the print run earlier in the day rather than banking the saved time as a later finish.
What to Publish, and What to Stop Publishing
Turkish distance-selling rules give you up to thirty days to perform the contract. Your customer gives you two. The gap between the legal maximum and the commercial reality is why so many stores publish a handling time that protects them legally and costs them commercially.
"Ships in 1–3 business days" is the classic example. It is written to cover the worst case, it tells the customer nothing, and it quietly re-prices your product against a competitor who wrote "dispatched today if you order before 14:00." One of those is a promise. The other is a disclaimer with a range attached.
Three things belong on the site, and each should be specific enough that a customer could catch you breaking it.
A cutoff time, with a weekend rule. Not "same-day dispatch" — a clock time, and what happens on Friday afternoon and over the weekend.
Handling time stated separately from transit time. They are different promises made by different parties and mixing them into one number is how a store ends up apologizing for a carrier and getting blamed for a warehouse. The shipping policy guide has the full structure.
A dispatch confirmation that means dispatch. If your notification fires at label creation, either rename it or move it. Both are cheap. Doing neither is what turns a good operation into a store that "never tells you anything."
Fix it: read your own shipping page as a customer who ordered at 16:00 on a Friday. If it does not tell them what happens next, it is not a shipping page, it is a legal disclaimer.
Five Numbers That Make Dispatch Time Manageable
You do not need a dashboard. You need five numbers, reviewed monthly, in whatever tool you already use.
Median order-to-label time. How long from payment to a printed label. Track the median and the ninetieth percentile — the second number is where your angry customers live.
Median label-to-first-scan time. The invisible half. This is the number that most often turns out to be the larger of the two, and the one nobody has ever seen.
Same-day hand-off rate, by weekday. The percentage of orders received before your cutoff that physically entered the carrier network the same day. Split by weekday, or Friday will hide inside the average.
Cutoff adherence. Of the orders that missed same-day hand-off, how many arrived before your published cutoff. Every one of those is a promise you broke in writing.
Support contacts per hundred orders, split by dispatch lag. The one that pays for the work. Group orders into lag bands and count the contacts each band generated. The relationship is usually steep, and it converts an operations metric into a cost you can put in front of anyone.
These sit alongside the delivery-side measures in the shipping KPI guide. The distinction worth keeping: those measure whether your carrier is doing its job, these measure whether you are doing yours.
Where a Shipping Platform Compresses the Clock — and Where It Can't
Dispatch time is made of three kinds of work: physical work, decisions, and typing. A shipping platform is very good at two of them and useless at the third. Being clear about which is which is the difference between a real improvement and a disappointment.
Here is what moves when orders, carriers, labels and tracking sit in one place, as they do in Shipink:
Orders arrive as they are placed. Every connected sales channel imports automatically, so the clock starts when the customer pays rather than when someone runs an export.
The carrier decision happens at the label, not in your head. Rates across 15+ carriers are compared at the moment of shipping, so nobody looks anything up.
Rules replace the repeated judgment call. Automation rules assign carrier and service by destination, weight and channel on Pro and Enterprise plans, which removes the decision from the critical path entirely.
Addresses get corrected before printing. AI address correction, also on Pro and Enterprise, catches the problems that otherwise surface at the label step and cost you a day.
Labels print in batches. Select the day's orders, print once, hand over once.
Both timestamps live in the same place. Label creation and the carrier's own tracking history sit on the same shipment record, which is what makes the label-to-first-scan gap measurable at all — along with per-carrier delivery time and success-rate reporting.
The customer hears from the real status. Notifications on email, and on SMS and WhatsApp for Pro and Enterprise, fire from carrier status changes rather than from your click, so a shipped message means the parcel moved. Delay alerts on Pro and Enterprise flag the shipments that have stopped.
And the honest limit. No platform picks, packs, or tapes a box. None of them makes the courier arrive earlier, opens the branch on Sunday, or moves the transfer vehicle. Shipink does not book carrier pickups for you — the hand-off remains a thing you schedule with your carrier. What the software removes is the decision-making and the re-typing wrapped around the physical work. In most stores at this size that is the larger half, but it is not the whole thing, and anyone telling you otherwise is selling you a disappointment.
What Not to Do
The failure modes here are consistent, and most of them come from treating a measurement problem as a capacity problem.
Don't fix dispatch time by promising less. Widening "1–2 days" to "1–5 days" makes the metric look better and the business worse. Fix the hours first, then re-promise from the new number.
Don't hire before you have removed the manual touches. A second person halves the hours but not the work. If three of the five steps before the label are typing, adding a typist is the expensive version of the fix.
Don't move your cutoff later to look competitive. A published 17:00 cutoff with a 14:00 pickup deadline behind it is not a faster store, it is a store that breaks a written promise every afternoon.
Don't compare your dispatch lag to a marketplace's. Their fulfillment model is different, their volume buys different pickup arrangements, and the comparison tells you nothing actionable. Compare yourself to your own median from last month.
Don't send "shipped" when you mean "label created." Everything else in this guide is optional. This one is just accurate labeling, and it prevents a specific, expensive, entirely self-inflicted category of support contact.
Don't change any of this after mid-October. You need roughly a month of live volume before you can trust that a change helped. Anything altered inside the last thirty days before peak is a change you will be debugging during the worst possible week — the peak season guide makes the case for that freeze in full.
Why September Is the Window
Two reasons, and they point at the same date.
The first is arithmetic. Early September is roughly ten weeks from 11.11 and a little over twelve from Efsane Cuma. Measuring a baseline takes a couple of weeks of normal volume, changing something takes a week, and confirming the change worked takes a month of live orders. Start in September and the measuring and the changing both land well before the mid-October freeze, with the confirmation month running right up to peak. Start in October and you are choosing between an unmeasured change and no change.
The second is that dispatch time is the first thing peak season breaks, and the only part of the delivery chain you can improve without anyone's permission. Adding a carrier is a procurement project. Negotiating a rate takes weeks and a volume history. Replatforming is a quarter of your life. Cutting your own dispatch lag needs no counterparty at all — it is your process, your cutoff, your notification text, and it is available this week.
There is also a compounding effect worth naming. Every hour you take out of dispatch is an hour of buffer against the carrier's peak-season slowdown. Transit times stretch in late November whatever you do. A store dispatching in six hours absorbs that; a store dispatching in thirty hours turns a two-day carrier delay into a week the customer will describe in a review.
The Dispatch Time Checklist
Measure first, before you change anything
- Median order-to-label time, computed from thirty consecutive shipments
- Median label-to-first-scan time, from the same thirty
- Both medians split by day of the week
- Both medians split by carrier
- The reason logged, for one week, on every parcel that missed same-day hand-off
Fix the hours
- Order import from every sales channel is automatic, not exported by hand
- Carrier selection is a rule, not a per-order decision
- Addresses are validated at order time, not at label time
- Invoices are issued automatically rather than re-typed
- Labels print in batches, and the print run has moved earlier in the day
Fix the cutoff
- The carrier's same-day pickup deadline is confirmed with your rep, in writing
- The branch's last outbound transfer time is known
- The published cutoff is the earliest of the three clocks, minus a buffer
- Friday has its own cutoff, and it is earlier
- A weekend hand-off has been priced and either bought or consciously declined
Fix what the customer sees
- The dispatch notification fires at hand-off, or its wording says what is actually true
- Handling time and transit time are stated separately on the site
- The cutoff time and the weekend rule are published, with clock times
- Support contacts per hundred orders are being tracked against dispatch lag
- Nothing in this list is being changed after mid-October
Stop Losing Days You Never Charged For
The uncomfortable thing about dispatch time is that it is entirely yours. There is no carrier to blame, no rate to renegotiate, no integration to build. The hours are being lost inside a building you control, to steps that are mostly administrative, and the customer is billing you for every one of them in patience they no longer extend.
The comforting thing is the same fact from the other side. Nobody has to agree to let you fix this. Measure two timestamps, find your real cutoff, move the print run earlier, and stop announcing shipments that have not shipped. Most stores find a day in there — a whole day, on every order, without touching the carrier or the warehouse.
Shipink brings the orders, the carriers, the labels, the notifications and both of those timestamps into one place, so the hours before hand-off stop being invisible. Start free and measure your own gap first — that number is usually the argument. If you would rather have someone look at it with you, get in touch.
