In August, everything works.
A homeware store ships around 70 orders a day. One person handles it: exports the orders, prepares the parcels, arranges the pickup, pastes tracking numbers back into the store, answers the handful of "where is it?" messages that come in. It takes a bit over three hours and it has never failed. There is no reason to change anything.
In late November that same store will do 250 orders on its worst day. Nothing about the process will have changed — which is exactly why it will fall apart. Three hours becomes twelve. The five daily "where is it?" messages become thirty. The one carrier that has always been fine will slip by two days in the third week of the month, and nobody will notice for four days, because nobody was measuring transit time in the first place.
This is the part that surprises people: peak season very rarely introduces a new problem. It takes the problems you have already decided to live with — the manual step, the unmeasured carrier, the single point of failure who also happens to be the founder — and multiplies them by whatever your peak multiple is. A process with one weak point at 70 orders a day has the same one weak point at 250. It just can't absorb it anymore.
The good news is that peak season is the most predictable crisis in e-commerce. You know the dates. You know roughly what your multiple will be. And unlike almost every other operational emergency, you get a three-month warning.
So this guide treats peak season shipping preparation as a schedule rather than a scramble: what to fix at 90 days out, what to rehearse at 60, what to freeze at 30, what to watch during the season, and what arrives in January that almost nobody budgets for.
One note before the plan: the dates below use the 2026 Turkish campaign calendar as the example. The T-minus structure is the part that matters — it works for any year and any peak, whether yours is Kasım (Turkey's November campaign month), Ramazan, or your own anniversary sale.
Peak Season Doesn't Create New Problems. It Multiplies Them.
The reason this framing matters is that it changes what you should do with the next three months.
If peak season created new problems, preparation would mean adding capacity: more staff, more carriers, more stock, a bigger plan. That's how most peak-season advice reads, and it's written for businesses with the leverage to act on it. At 50–5,000 orders a month you don't have that leverage. You are not getting a dedicated capacity commitment from a national carrier. You are not opening a second warehouse in October.
What you do have is control over your own multipliers. Every one of these is something you own:
- Manual touches per order. At three minutes of handling per order, 70 orders is three and a half hours and 250 orders is more than twelve. The minutes don't scale — your day does not get longer.
- Failure rates you can't see. A 2% address-problem rate is roughly one parcel a day at 70 orders. At 250 it's five a day, every day, for six weeks — and each one becomes a support conversation, possibly a redelivery, possibly a return.
- Single points of failure. One carrier, one printer, one person who knows the process, one login. Peak season is a stress test that finds all four.
- Blind spots. If you don't know your normal on-time rate, you cannot detect the week it drops. You will find out from customers instead, which is the most expensive way to learn anything.
That's the entire premise of the plan below: spend the runway reducing multipliers, not adding capacity. Reducing a multiplier in September pays out every single day of the season.
What Actually Happens to Parcels When Volume Spikes
It's worth being specific about what the carrier network does under load, because "it gets busy" is not something you can plan against.
Transit times stretch, and not by a little. Sendcloud's peak-season analysis puts the surge on Cyber Monday at over 200% above the average daily parcel volume, with average delivery times in December rising by up to two days. Two days is not a rounding error when your product page promises delivery in two to three.
Loss and damage rates rise faster than volume. Looking at Black Friday/Cyber Monday 2024, Sendcloud recorded parcel losses up 20% and damaged parcels up 34% versus normal periods, with some markets much worse — parcel losses spiked over 40% in the Netherlands and damages jumped 37% in France on Cyber Monday. Sorting under pressure is rougher sorting. Packaging that survives a normal week doesn't necessarily survive a peak week.
The Turkish network is genuinely absorbing more each year, but it absorbs it at the edge of capacity. Sürat Kargo's own figures make the shape of this visible: a daily record of 750,000 parcels in November 2024, then 900,000 parcels in a single day in November 2025, and around 11.5 million parcels across that November. They got there with automated sorter investments, extra staff and vehicles, extra shifts and weekend working. That's what "the network is ready" actually looks like from the inside — and it's why the third week of the month behaves differently from the first.
Exceptions grow with volume, and they grow at your expense. If 1% of your shipments hit an exception — wrong address, failed attempt, damage, missing scan — then 2,000 parcels a month means 20 exceptions and 8,000 means 80. The work per exception doesn't shrink. This is the number most stores discover too late, because failed deliveries and lost or damaged parcels are handled as one-off annoyances rather than a rate you can forecast.
The honest summary: during peak season your carrier will be slower and slightly less careful than it is today, and neither of those is something you can negotiate away in November. What you can do is stop being surprised by it, and make sure a slow week is visible to you within a day rather than two weeks later.
In Turkey, Peak Season Is a Month — Not a Weekend
Most peak-season content is written around a four-day event: Black Friday to Cyber Monday, prepare, survive, recover. That's not what the Turkish calendar looks like. Kasım is a full month of stacked campaign days, and the operational consequence is that you're not preparing for a sprint. You're preparing for four to six weeks of sustained above-normal volume with no recovery gap in the middle.
Here's the 2026 shape, with a working T-minus reference:
| Date (2026) | Event | Countdown |
|---|---|---|
| 11 November | 11.11 / Singles' Day | Day zero for most stores |
| 24 November | Öğretmenler Günü (Teachers' Day) | Mid-season spike |
| 27 November | Efsane Cuma (Black Friday) | The peak day |
| 30 November | Cyber Monday | Highest label volume |
| December | Year-end and New Year gifting | Sustained, plus delivery-deadline pressure |
| Late Dec – Jan | The returns wave | The bill for a good November |
Work backwards from the earliest real spike rather than from Black Friday. If 11.11 is when your volume first jumps, then T-90 is roughly the second week of August, T-60 is mid-September, and T-30 is mid-October. That's the runway, and it's shorter than it feels. The practical window for readiness work is July through September; by October you want to be stabilizing rather than building, because anything you change in October is still unproven when the volume arrives. Everything in this guide is arranged around that.
One number worth planning against. The Ministry of Trade's Türkiye'de E-Ticaretin Görünümü 2025 report shows that in the November campaign month, e-commerce volume rose 41.5% year-on-year while the number of products and services sold rose 11.6%. Those two figures are far apart, and the gap is mostly price. It matters operationally because your capacity constraint is parcels, not turnover: if you plan warehouse hours, packaging stock, and carrier conversations around a 40% revenue jump, you will over-buy in some places and still be wrong. Plan capacity on parcel count, and use your own last-November order data — not your revenue growth — to set the multiple.
And one baseline expectation you're now competing against. The same report puts the average e-commerce delivery time in Turkey at 42.2 hours, down from 46 hours two years earlier, with 53% of orders delivered within 24–48 hours. Fast delivery stopped being a differentiator and became the floor. During peak season your customers do not lower that expectation because it's busy — they compare you to the marketplace parcel that arrived the next morning.
Step Zero: Take a Baseline You Can Compare Against
Do this first, this week, before you change anything else. It takes an afternoon and it is the highest-leverage hour in the whole plan.
You cannot manage peak season by feel. "It seems slower this week" is not actionable, and by the time it's obvious enough to be certain about, you've shipped four thousand parcels into a degrading service. So capture what normal looks like now, in August or September, while the network is calm.
Five numbers, per carrier, not blended:
- On-time rate against the date you promised the customer — not against the carrier's own service claim. These are different numbers and only one of them affects your reviews.
- Average transit time and its spread. The average tells you what to promise. The spread — how much your slowest 10% differs from your median — tells you how much buffer you need. Peak season widens the spread before it moves the average.
- First-attempt delivery rate. Every failed first attempt is a redelivery, a support message, and a candidate return.
- Exception rate. Anything that didn't move cleanly from pickup to delivery.
- True cost per delivered order. Not per label — per delivered order, including redeliveries, returns, re-invoiced weight differences, and surcharges.
If you can't produce these five per carrier in under ten minutes, that's your first finding, and it's a bigger one than any of the numbers would have been. Fixing the measurement gap is more urgent in August than fixing the carrier, because measurement is what tells you whether anything else you do in the next three months worked. The mechanics of all five, including how to calculate cost per delivered order properly, are in the guides on carrier performance metrics and shipping KPIs.
Write the five numbers down with the date. In the third week of November, that piece of paper is what turns "I think Aras is slow this week" into "our median transit on the İstanbul–İzmir lane went from 1.8 to 3.4 days, so we're moving the express SKUs."
T-90 to T-60: Change the Structural Things
This is the only window in which you should touch anything structural. Anything you change here still has a full month of real volume ahead of it before peak begins — which is the minimum needed to know whether it actually works.
Get a second carrier genuinely live, not merely signed
Not "we have an account with them." Live: shipping real orders, every day, in some meaningful share of your volume — 15–20% is enough. The reason is simple. A carrier you've never used is not a backup; it's an unvalidated integration plus an unknown pickup relationship plus a barcode format your labels have never printed. Discovering all three during the week you actually need them is how a bad week becomes a bad month.
A second carrier gives you three things at peak: a route around one network's regional slowdown, per-parcel choice between them, and negotiating position for next year. The multi-carrier strategy guide covers the split logic, and the Turkish carrier comparison is a reasonable starting point for who to add.
Remove the per-order manual touches
Go through your own process and count the times a human touches a single order: opening it, copying an address, choosing a carrier, entering a weight, printing, writing the tracking number back into the store, marking it shipped, telling the customer. Each of those is a fixed cost per order that peak season multiplies, and each is a place a typo enters your data.
The realistic target is that order import, rate selection, label creation, and tracking write-back all happen without anyone opening an individual order. Vendors across this category consistently report per-order handling dropping from around three minutes to roughly twenty seconds once that's in place. At 250 orders on your worst day, that's the difference between more than twelve hours of work and about ninety minutes. The full picture is in the shipping automation guide.
Fix it: pick the two most-repeated manual steps and eliminate those first. Don't attempt a total rebuild in September — you'll be mid-migration in October, which is precisely the situation the change freeze below exists to prevent.
Clean up address quality now
Roughly 2% of parcels have an address problem, and peak season is the worst possible time to be arguing about apartment numbers. Address validation at checkout, a required phone number, and correction rules applied before the label prints all cost you nothing per order once configured. Every one you catch in advance is a failed delivery that doesn't happen, and failed deliveries in a saturated network don't get retried quickly.
Settle packaging and weight before volume arrives
Two peak-specific reasons, beyond the usual cost argument. First, damage rates rise measurably under peak sorting loads, and the parcel that survived a quiet August is not automatically the parcel that survives late November. Second, weight and dimension errors get re-invoiced as desi differences, and those corrections land on the invoice after the season, when your margin is already spent. Standardize your box sizes and weigh accurately now — the method is in the packaging optimization guide.
Sort invoicing while volume is low
Whatever your invoicing process is, it is about to run four times as often. If e-fatura and e-arşiv issuing is a separate manual chore attached to each order, that chore is going to consume the exact hours you need for packing. Get it automated and issuing off the order data in September, not on 11 November.
T-60 to T-30: Rehearse Instead of Rebuilding
By mid-September the structural work should be done or nearly done. This window is for finding out whether it holds.
Run a peak-volume day on purpose
Pick your best sales day this window and treat it as a rehearsal: process the whole day's orders in one batch pass, with the process you intend to use in November, and time it. Then check the failure points rather than the total. Did every order that came in before your cutoff actually ship that day? Did the labels all print first time? Did the pickup happen when it was supposed to? Where did the queue form?
If you can't create the volume, simulate the batch: prepare labels for the day's orders in a single pass rather than one at a time. The point isn't the volume, it's discovering which step doesn't survive being done fifty times consecutively.
Set your daily cutoff time and mean it
The single most common cause of peak-season delay is not carrier capacity — it's the gap between when an order is ready and when the carrier physically collects it. A parcel that misses today's pickup does not lose an hour; it loses a day, and during peak season it can lose two, because tomorrow's pickup is already full.
So fix a cutoff time, set your internal deadline earlier than the carrier's, publish it, and hold it. Then verify the collection actually happens at that time in October, before it matters.
Decide what you're going to promise
Peak-season delivery promises are where stores manufacture their own January complaints. If your normal promise is two to three days and December transit stretches by up to two, your promise is now wrong for a month, and being wrong about a delivery date costs more than being slow — roughly 70% of customers won't order again after a bad delivery experience.
Add buffer to your displayed dates for the season, publish the campaign-period cutoffs for delivery before specific dates, and update your shipping policy to match. The mechanics of setting and displaying dates you can actually hold are in the estimated delivery dates guide. And if you're planning a free-shipping threshold for the campaign, do the margin arithmetic now — free shipping that works at August volumes and August carrier performance can quietly invert during a month with more redeliveries and more returns.
Plan for the cash side of cash on delivery
If a meaningful share of your orders are kapıda ödeme, peak season affects you twice: the share typically rises during campaign periods with new customers, and the collection cycle means a big November is also a month where a lot of your revenue is sitting in the carrier's system rather than your account. Failed COD deliveries also cost more than failed prepaid ones — you pay to send it and pay to get it back. Confirm your remittance timelines and reconciliation process before the volume arrives; the COD guide has the operational detail.
T-30 to Day Zero: Freeze, Then Communicate
Thirty days out, the posture changes completely. Stop improving. Start protecting.
The change freeze
From roughly T-30, don't change your shipping stack, your carrier mix, your label format, your warehouse layout, or your order-management flow. Not because those changes are bad — because you no longer have the runway to discover that one of them was.
This is the clearest answer to "when is it too late?" You need about a month of live volume to trust a change, and a month is what you have left. After that point, the correct response to a good idea is to write it on a list dated 15 January.
The one exception worth making: turning on something that fails safe. Enabling additional notifications, adding buffer to displayed delivery dates, or bringing a rehearsed carrier's share up is low-risk. Migrating a system is not.
Turn on every notification you have
This is the highest-return thing you can do in the last month, and it takes an afternoon. WISMO — "where is my order" — is 35–60% of all e-commerce support tickets, and each one costs somewhere between $5 and $8 to handle. During peak season both the volume and the cost per ticket rise, because your team is already stretched.
Proactive status notifications reduce WISMO contacts by 40–80%. That's not a customer-experience nicety in November; it's the difference between your support inbox being manageable and being abandoned. In Turkey the channel choice is easy: WhatsApp is the most-used application in the country, reaching 88.6% of individuals according to TÜİK's 2025 data, and it costs a fraction of SMS. The channel economics and template mechanics are covered in the WhatsApp notifications guide and the broader WISMO deflection guide.
Pair it with a tracking page customers can actually check themselves. A branded tracking page that shows real status is a support agent that works at 2am during a campaign week.
Prepare the support side for volume, not just for questions
Write the macros now: delayed parcel, failed attempt, damaged item, wrong address, "can I still get it by the 31st?". Decide your policy in advance for the parcel that's four days late — replace, refund, or wait — because deciding it case-by-case at 60 tickets a day is how response times collapse. And make sure at least two people can run the entire shipping process end to end, including the login.
Check the boring physical things
Thermal label stock and a spare roll. A backup printer or a tested fallback. Packaging materials for your peak multiple, ordered in October, because your supplier has a peak season too. It's a fifteen-minute list and it prevents the stupidest possible failure mode: a fully automated shipping operation stopped by an empty label roll on 27 November.
During Peak: The Ten-Minute Daily Check
Once the season starts, your job stops being improvement and becomes detection. Every morning, look at four things:
- Yesterday's orders that didn't ship. The number should be near zero. If it isn't, you have a throughput problem today, not tomorrow.
- Median transit time per carrier, per major route, versus your baseline. This is what the September baseline was for. A carrier drifting from 1.8 to 3.4 days on your busiest lane is a decision, and it's only a decision if you see it in days rather than weeks.
- Exceptions and failed attempts, as a count. Rising exception counts are the earliest warning that a region is congested.
- WISMO ticket volume. A spike here usually means the tracking or notification flow has broken somewhere, not that customers suddenly got impatient.
Set two rules in advance so you're not making judgment calls while exhausted. Something like: if a carrier's median transit on a lane exceeds baseline plus one day for two consecutive days, shift new orders on that lane to the other carrier. And: if unshipped orders exceed one hour of packing capacity, the cutoff moves earlier today. Written rules beat in-the-moment decisions in week four of a campaign month.
After Peak: The Two Bills Nobody Budgets For
Peak season is not over on 1 December. Two things arrive afterwards, and both of them take back part of what November earned.
The returns wave
January is the year's real returns peak. Global data for the 2025–26 holiday season showed customers returning around 12.2% of online orders in early January, and category rates run far higher — apparel and electronics are the extremes. The better your November, the heavier your January.
If your returns process is manual — customer messages you, you reply with an address, you wait for a parcel you can't identify, you refund when you find it — then multiply that by your November. Prepare it while you still have attention to spare: a self-service return request, return labels you can issue without a conversation, and a status the customer can see. The returns management guide has the full flow. Do this in October, not on 3 January.
The carrier invoice
The second bill is quieter. Weight and dimension corrections, redelivery charges, additional-service fees, address-correction charges and COD service fees mostly arrive on invoices after the season, covering shipments you priced in November. Stores that don't reconcile simply absorb the difference and conclude that peak season was less profitable than it looked.
So plan a reconciliation pass for mid-January: compare what you were charged against what you shipped, per carrier, and specifically look for re-invoiced desi differences and repeated surcharges on the same lane. Two useful outcomes come out of it — a corrected view of your true cost per delivered order, and the evidence you need for a better rate next year. Take that data straight into your rate negotiation, because "here is our real volume and here is where you missed the service level" is a much stronger position in February than in November. The wider cost picture is in reducing shipping costs.
Running the Whole Plan on One System
Most of the plan above is process, and you can do a lot of it with discipline and a spreadsheet. But the parts that carry the season — batch label creation, per-carrier visibility, automatic notifications, one queue across every channel — stop being feasible by hand at exactly the volumes peak season produces. That's the gap Shipink is built for.
Concretely, for a store getting ready for Kasım:
- Connect every sales channel you sell on — Shopify, WooCommerce, ikas, Ticimax, IdeaSoft, Shopier, PrestaShop, OpenCart, Wix, Etsy, Trendyol — so campaign orders land in one queue instead of five browser tabs. This is the single biggest change for anyone selling in more than one place; the reasoning is in the multichannel order management guide.
- Bring your own carrier agreements, or use Shipink's — including kapıda ödeme — and run more than one carrier account on the same channel, which is what makes the T-90 "second carrier live" step practical.
- Compare rates automatically across 15+ carriers per parcel, so the carrier choice is a priced decision rather than a habit.
- Create and print labels in bulk in one pass, which is the step that turns a twelve-hour peak day into a ninety-minute one.
- Set automation rules — route by weight, destination, channel, payment type or value — so peak volume doesn't mean peak decisions.
- Send automatic status notifications over WhatsApp, SMS and email from a branded tracking page, which is what keeps November's support load survivable. WhatsApp and SMS notifications are included on Pro and Enterprise plans.
- Issue e-fatura and e-arşiv automatically, and read per-carrier performance reporting for the daily ten-minute check and the January reconciliation.
If you're evaluating this against alternatives — including doing nothing — the buyer's guide sets out what to actually test. And test it now: the whole argument of this guide is that a system you adopt in August is trustworthy by November, while a system you adopt in November is a second problem on top of the first one.
What Not to Change This Close to Peak
Restraint is part of readiness. A few things you should specifically not do:
Don't replatform. Whatever is wrong with your storefront, October is not when you fix it. A platform migration during peak-season prep risks the entire season to solve a problem you've lived with all year.
Don't add a carrier you can't measure. A third carrier you have no performance data on doesn't reduce your risk — it adds a variable you can't see. If you can't read its on-time and exception rates separately from the others, it isn't a backup; it's a coin flip.
Don't chase the cheapest rate into a slower service. During peak, the gap between carriers is service reliability, not price. Saving a little per parcel and losing two days of transit in the week before Efsane Cuma is a bad trade — the redeliveries, tickets, and returns cost more than the saving.
Don't hire your way out of a manual process. A second person halves the hours but not the work: you're still paying for three minutes of handling per order, now at two people's cost, plus training time you don't have in November, plus a second person who can make the same typo. Remove the step first, then add people if you still need them.
Don't promise a date you haven't held in October. Promising Efsane Cuma delivery by a specific day is a genuine conversion lever, but only if your baseline data says you can. Promise the date your slowest 10% can hit, not the one your median can.
Your Peak Season Readiness Checklist
Now (Step Zero)
- Baseline recorded, per carrier: on-time vs promised date, median transit + spread, first-attempt rate, exception rate, true cost per delivered order
- Last November's order data pulled to set your peak multiple — on parcel count, not revenue
- Your own T-90 / T-60 / T-30 dates written down, counting back from your earliest spike
T-90 to T-60 (structural)
- Second carrier live with real daily volume, not just an account
- Two most-repeated manual steps per order eliminated
- Address validation and required phone number at checkout
- Box sizes standardized, weights accurate, desi checked
- e-Fatura / e-arşiv issuing automated
T-60 to T-30 (rehearse)
- One full day processed as a single batch pass, timed
- Daily cutoff time set, published, and internal deadline set earlier
- Pickup reliability verified in the real world
- Displayed delivery dates buffered for the season; shipping policy updated
- Campaign free-shipping threshold checked against peak-season margin
- COD remittance and reconciliation process confirmed
T-30 to day zero (freeze and protect)
- Change freeze in effect; ideas parked on a January list
- WhatsApp / SMS / email status notifications live
- Branded tracking page live and linked from order confirmations
- Support macros written; late-parcel policy decided in advance
- At least two people can run shipping end to end
- Label stock, spare roll, backup printer, packaging ordered
During and after
- Ten-minute daily check running, with two written switch rules
- Returns process ready before December, not during January
- Mid-January carrier invoice reconciliation booked
- Post-season notes written for next year's baseline
Start Now, Because the Date Doesn't Move
Almost every operational problem in e-commerce arrives without warning. This one sends a calendar invite three months ahead.
The stores that come out of Kasım with better numbers than August aren't the ones that worked hardest in November. They're the ones that spent August and September removing the manual steps and the blind spots, spent October rehearsing and then deliberately stopped changing things, and treated the season itself as a monitoring job rather than a heroic effort.
The plan is not complicated. Take the baseline this week. Fix the structural things while you still have a month of runway behind each change. Rehearse in October. Freeze at thirty days. Watch four numbers a day. Book the January reconciliation now, while you still remember why it matters.
And do the first part now rather than in October — not because October is too late to care, but because it's too late to change. A shipping process you trust in September is an asset in November. One you're still learning is a liability.
Start your free Shipink trial and get your peak-season setup running while you still have the runway to test it.
