Growth
The peak season plan: what to lock by October
A dated checklist for Q4: feed and account health, promotions, creative, email and site readiness, plus the freeze dates that keep peak trading calm.
By CartKernel · Published
The stores that trade calmly through November and December are the ones that finished the work in October. Peak season rewards preparation and punishes improvisation, because every fix costs more when traffic is at its highest and every approval queue is at its longest.
This is a dated plan. It assumes a Northern Hemisphere Q4 with promotional weekends in late November and a shipping deadline in the third week of December. Shift the dates to your own calendar, and keep the sequence.
By the end of September: fix the foundations
Clear every account and feed error. Disapprovals, warnings and account notifications should be at zero before October, not because the errors are urgent in September but because the review and resubmission cycles are unpredictable in November. Read the account guidelines properly once, and check the areas that commonly trip stores up: price and availability accuracy, shipping and return policy information, and any category with its own requirements.
Confirm the feed refresh cadence. Availability and price change fastest in peak. A feed that updates once a day will show items as available hours after they sold out, and mismatches between the feed and the site are the most common cause of disapproval at exactly the wrong moment. How often should a product feed update covers the mechanisms, and if fetches have ever failed, Merchant Center feed fetch failed is the fix list.
Label the catalogue for the season. Custom labels for seasonal relevance, stock depth, margin tier and gift suitability need to exist before you want to bid on them. Adding custom labels in November means restructuring campaigns during the period you least want to reset learning.
Check payment and billing on ad accounts. A declined card on the biggest spending day of the year is an avoidable outage. Raise credit limits, confirm billing thresholds, and give a second person access.
Verify tracking end to end. Place a test order on mobile and desktop and confirm the purchase reaches every platform once, with the right value. Peak is not the time to discover a measurement gap, because every optimisation decision for the next ten weeks depends on it.
Early October: content and creative
Build the gift and seasonal landing pages now. New collection pages need time to be crawled, indexed and to accumulate internal links before they can earn organic traffic in November. A gift guide published on 20 November is a paid landing page, not an organic asset. Publish in early October, link from the main navigation, and let them settle.
Where the same pages ran last year, keep the URLs rather than creating new ones. A page that has been earning since last November restarts from zero if it gets a new address.
Produce creative in batches. Assemble the full set now: offer variants, product-led versions, gifting angles, and formats for each placement. Producing creative during peak means producing it late and shipping whatever is ready rather than what tested well.
Write the email and SMS calendar. Every send from mid-October through early January, with dates, audiences and offers. Write the sends themselves in October where you can. The calendar’s purpose is to stop the daily debate about whether to send, which is what produces both over-sending and missed days.
Warm the list. Increase send frequency gradually from early October rather than jumping from weekly to daily on the promotional weekend. Sudden volume increases on a list with cold segments cause deliverability problems at the moment they cost most. How often should an online store email its list covers the cadence question outside peak.
Mid October: offers, stock and shipping
Decide the promotional structure and stop revising it. Write down for each promotional period: the offer, the products included and excluded, the start and end times in a stated time zone, the margin impact, and who can approve a change. An offer decided a week before is an offer that has not been tested against margin and cannot be built into creative in time.
Where you advertise on shopping surfaces, promotions have their own data requirements and their own review process. Submit them well ahead of the start date, because a promotion approved after its start is a promotion that did not run. Are Merchant Center promotions worth using covers whether they earn their place for your category.
Set the shipping deadlines and publish them. Get carrier cutoff dates, add your own handling time, subtract a safety margin, and publish the resulting dates on the site by mid October. Deadlines are a conversion tool in December and a support cost if they are wrong.
Extend the return window and say so. Gifts bought in November are opened in December. A return window that expires before the gift is unwrapped generates complaints regardless of the policy’s technical correctness. Publish the extended window everywhere the policy appears, including the advertising account settings.
Plan for stock-outs before they happen. Decide now what a product page does when it sells out: stay live with a back-in-stock signup and alternatives, or redirect. Decide what campaigns do when a bestseller runs out, so budget moves rather than sitting on unavailable products. Out of stock page handling sets out the options.
Late October: the site itself
Test at peak load. Whatever your infrastructure, know what happens when traffic multiplies. Test the checkout under load, not just the homepage.
Measure the templates, not the homepage. Product and collection pages carry the traffic. Check Core Web Vitals on a mid-range phone against a real product page, and use the Core Web Vitals checker for a quick read on the templates that matter.
Audit installed apps and scripts. Every seasonal widget, banner app and tracking tag added over the year costs load time on every page. Remove what is not earning. This is the cheapest speed work available and it takes an afternoon.
Walk the whole purchase path on a phone. Home, category, filter, product, add to cart, checkout, payment, confirmation. Then do it as a gift purchase to a different address. Most peak problems are found this way rather than in a report.
Check the abandoned cart and post-purchase flows. Timings that work in a quiet month can be wrong in peak, when shoppers are comparing across several stores in an evening. Abandoned cart flows are the highest-value automation running during peak, and they should be reviewed rather than assumed.
1 November: the freeze
From the start of November, the rules change.
Code freeze. No theme changes, no app installs, no checkout modifications, no platform upgrades except security fixes. Every change is a risk taken during the period when a mistake is most expensive.
Campaign structure freeze. Budgets, bids and creative rotation continue. Restructuring, new campaign types and bidding strategy changes stop. Automated bidding needs stable inputs, and a strategy changed in mid November spends peak in a learning period.
A named person on call, with a rollback plan. Write down who is called, what they can revert, and how. This document takes twenty minutes and is the difference between a ten-minute incident and a lost evening.
A daily check, not a dashboard vigil. Each morning: orders against yesterday and against the same day last year, feed and account status, ad spend pacing, stock on the top twenty products, site response, and the support queue for repeated complaints. Fifteen minutes, same time, written down.
During peak: what to actually change
Only three things should move day to day.
- Budget, toward what is working. Watch impression share on your best campaigns. Losing impression share to budget during peak is money left on the table.
- Product availability in campaigns. Pause or exclude products that have sold out so spend follows what you can ship.
- Offer messaging in email and on site. Copy and banner changes within the structure already built.
Everything else waits. The plan was made when there was time to think.
After the peak: December and January
Publish the last order dates prominently and then change the message. Once shipping deadlines pass, switch to digital gift cards and local pickup where you offer them. A store still advertising delivery by the 24th on the 23rd is generating refunds.
Prepare for returns. January volume is a fulfilment and a merchandising event. Capture reasons at scale and use them, because the returns data from peak is the most complete picture of product page accuracy you will get all year.
Run the retention sequence. A first-time buyer acquired in November is a new customer, not a completed order. The second purchase in the first quarter is what turns a busy season into a base, and that sequence should have been written in October along with everything else.
Write the debrief in the first week of January. What sold, what ran out, what the support queue was about, which offers worked, what broke. Write it while it is fresh, and it becomes next year’s plan.