Will Your Site Survive Black Friday Traffic?

A store that runs comfortably through October can fall over at four times the traffic, and the failure rarely looks like the site going down. It looks like checkout timing out for one in twenty people on the busiest evening of the year.

Black Friday is on 27 November and the code freeze lands next week, in the week commencing 9 November. That makes this the last week to test properly and still have time to fix what the test finds. This post covers what to check, what breaks first, and what to do when something fails on the day.

Why does peak traffic break sites that work fine in October?

Because the load arrives in a spike rather than a curve, and it concentrates on the slowest parts of the stack.

Normal trading spreads visitors across the day. A promotional launch puts them through the same three pages in the same ten minutes, which is a different engineering problem. Everything that queries inventory, applies a discount or writes an order hits at once, and those are the slowest operations you have.

Third party scripts make it worse. Every pixel, review widget, chat tool and personalisation script adds a request, and under load they degrade before your own code does.

What should you test this week?

The journey rather than the homepage. Most load testing checks the landing page, which is the part least likely to fail.

Test the full path: collection page, product page, add to basket, basket, checkout, payment, confirmation. Run it on mobile on a mid range device over a normal mobile connection rather than on office wifi on a new laptop, because that is what most of your peak traffic actually is.

Then test it with the promotion applied. A discount code, a bundle rule and a free shipping threshold all add logic to the basket, and the version of your site you tested in October was not running any of them.

What breaks first?

In rough order: third party scripts, then the basket, then inventory, then payment.

Scripts time out and block rendering, so the page appears broken while the server is fine. The basket is next, because discount logic and cross sell rules are the heaviest thing on the page. Inventory follows, particularly where stock syncs between a warehouse system and the store. Payment is usually the most robust, since the providers are built for this, though a payment timeout is the most expensive failure of the four.

The useful exercise is deciding in advance which of these you would switch off to keep the rest working. A review widget is worth sacrificing to keep checkout alive, and knowing that beforehand turns a crisis into a decision.

How fast does the site need to be?

Fast enough that the Core Web Vitals thresholds hold under load, not just in a quiet test. Largest Contentful Paint at 2.5 seconds or under, Interaction to Next Paint at 200 milliseconds or under, and Cumulative Layout Shift at 0.1 or under.

Interaction to Next Paint is the one that matters most during a promotion, because it measures how long the page takes to respond when someone taps. A site that renders quickly and then ignores the first three taps on the add to basket button feels broken to the person using it, whatever the page speed score says.

Measure on real traffic rather than in a lab tool. Field data from your own visitors tells you what is happening on the devices and connections your customers actually have.

What should you freeze, and when?

Non essential code changes from next week, in the week commencing 9 November, which is two weeks out from Black Friday.

That covers theme changes, new apps, script additions and anything touching the basket or checkout. The reason is not superstition. It is that a bug introduced on 20 November has four days of real traffic to be discovered in, and the people who could fix it are the people already handling everything else.

Content changes stay available: merchandising, copy, banners and product data all need to move during peak. The full sequence of what locks when is in the peak season freeze calendar.

What is your rollback plan?

A tested way to get back to a known good version in minutes, written down, with the person who can do it named.

Three things make this work. A copy of the current theme saved before the freeze, so there is something to roll back to. A documented order for disabling third party scripts, so you can shed load without guessing. And a way to see the site as a customer does, since internal tools often cache past a failure that visitors are hitting.

Test the rollback this week rather than assuming it works. A backup nobody has restored is a hope rather than a plan.

Who is watching over the weekend?

Someone named, with access, for each shift across the Cyber 5 window from 26 to 30 November.

Set up alerting that reaches a phone rather than an inbox: checkout error rate, payment failures, server response time and conversion rate against the same hour the day before. Conversion rate is the broadest signal you have, because almost every technical failure shows up there first.

Agree the threshold that triggers action and the person who decides, which sits alongside the wider trading decisions in peak escalation rules.

What this protects

The most expensive traffic of the year. Portfolio cost per acquisition ran $33.01 across Cyber 5 and $50.16 by mid December, so a checkout failure during peak is lost margin on traffic you have already paid for at the top of the market.

The commercial frame for peak, including the maximum acquisition cost to hold through November, is in the Q4 Profit Playbook, free and written for founder led DTC brands on Shopify at $5M to $30M.

The conversion side of the same work is in checkout optimisation, and Webtopia runs conversion alongside acquisition for founder led brands as an ecommerce marketing agency.

Frequently asked questions

How much traffic should I test my site for?

Test well above your expected peak hour rather than your expected peak day, since promotional launches concentrate visitors into minutes rather than spreading them across hours.

When should I freeze site code before Black Friday?

Two weeks out, in the week commencing 9 November 2026. Content and merchandising stay editable; theme, app and checkout changes stop.

What are the Core Web Vitals thresholds?

Largest Contentful Paint at 2.5 seconds or under, Interaction to Next Paint at 200 milliseconds or under, and Cumulative Layout Shift at 0.1 or under, measured on real visitor data.

What usually fails first on Black Friday?

Third party scripts, then the basket, then inventory syncing, then payment. Decide in advance which ones you would disable to keep checkout working.

Do I need a rollback plan if I have a backup?

A backup you have never restored is untested. Save a known good theme version before the freeze, document the order for disabling scripts, and rehearse both this week.