Run a Black Friday pre-mortem before the traffic shows up

Pretend it's Saturday and Black Friday went wrong. Write down why today, then spend the quiet week fixing each line.

Sitanshu 6 min read ·
An empty two-lane road running straight through flat farmland under a pale, hazy sky, a guardrail on the left
Photo: Unsplash

A Black Friday pre-mortem is one short meeting where you assume the weekend already went wrong, then write down why. You walk out with a list of failures while there’s still time to prevent them.

Black Friday is November 27. That’s 13 days.

The method comes from Gary Klein, who wrote it up in Harvard Business Review in 2007. The team pictures a project that “has just failed” and works backward to the reasons.

The wording does the work. “What could go wrong?” invites polite answers. “Why did it go wrong?” invites specific ones.

So here’s ours, for an imagined Shopify store. None of it happened. All of it can.

A hand-drawn incident report on a clipboard, titled Black Friday, November 27, with a pencil stamp reading imagined. Five timed lines: 6:02 am, email links to a hidden page. 9:40 am, gift-note box gone at checkout. 11:15 am, quick fix on the live theme. 12:01 pm, phones crawl at the peak. 2:30 pm, a vendor we forgot goes down. A pencil note at the bottom says none of it happened, all of it can.
Our imagined report. Every line below is one section, with a check you can run this week.

6:02 a.m. The early-access email opens a page nobody can see

The email was scheduled last week. The collection it links to goes live at 9.

For three hours, your most loyal customers tap through to a page that isn’t there yet.

The same goes for a discount code that starts later than the email announcing it, or an SMS that points at last year’s sale.

Check this week: open every scheduled email and text on your phone, signed out, and tap every link. Then put each discount’s start time next to the send time of the message that announces it.

9:40 a.m. The gift-note box quietly disappears from checkout

On November 6, Shopify started showing app developers how often their checkout extensions fail to load, and on which stores (Shopify’s changelog). Extensions are the add-ons inside checkout: gift notes, delivery dates, upsells.

When one fails, Shopify says, buyers don’t see the experience the app built.

The shopper never knew it was supposed to be there. So nobody reports it.

Two hand-drawn phone checkouts side by side, labeled example. The left phone, titled what you built, shows contact, delivery, a boxed gift note field reading happy birthday, and a black pay now button. The right phone, titled what a shopper saw, shows the same screen with a dashed empty gap where the gift note box should be. A pencil note underneath says nobody reports a missing box.
Example checkout. When an extension fails to load, buyers don’t see it, per Shopify’s changelog.

Ratings and admin alerts for store owners are “coming soon,” per the same note: a warning when more than 5% of loads fail for three days, and critical above 50% for two. Soon may not mean before Black Friday.

Check this week: list every app that adds something to your checkout. Email each vendor one question: what’s our store’s mount failure rate in your Dev Dashboard this month? If you’d rather have someone read those logs for you, that’s the work we do.

11:15 a.m. Someone makes a “quick fix” on the live theme

The banner has a typo. Someone opens the editor, fixes it on the live theme, and a section slips out of place on phones.

We suggested freezing your store on November 13. A freeze only holds if one named person can say no, and a way back is ready.

Shopify lets you duplicate a theme: in Online Store, open the theme’s ⋯ menu and click Duplicate. Keep a copy of today’s live theme, and publishing that copy puts the design back the way it was.

A theme holds your design, not your products or prices, so rolling back won’t undo a price change.

Today: duplicate your live theme and rename it “Backup Nov 14”. Rolling back becomes one publish.

12:01 p.m. The busiest minute arrives, and phones feel it first

Last Black Friday, sales across Shopify peaked at 12:01 p.m. Eastern, at $5.1 million a minute, per Shopify. Shopify plans for that minute. Your theme, apps and tracking scripts are on you.

And most of the visits come from a phone. Adobe forecasts a record 63% of Thanksgiving online spend will come through mobile devices, against 57.4% for the whole season.

A hand-drawn bar chart titled Adobe's forecast, US online spend in billions of dollars, labeled forecast. Thanksgiving 6.9, Black Friday 12.9, Cyber Monday 15.1, bars drawn to scale from zero. A pencil note above the Thanksgiving bar says 63% on phones.
Adobe’s 2026 holiday forecast for US online spend. Forecasts, not results.

Check this week: on your own phone, over cellular rather than office Wi-Fi, load the homepage, your best seller and checkout. Then buy something real and refund it. Anything that feels slow goes on the list to fix or switch off before the 27th.

2:30 p.m. A vendor you forgot you use goes down

On November 18, 2025, Cloudflare’s network started failing at 11:20 UTC, and visitors to sites that rely on it saw error pages. Core traffic was largely back by 14:30 UTC. Cloudflare’s own write-up traced it to a database permissions change, not an attack.

The lesson isn’t about one company. Your store leans on a chain of other people’s systems, and any link can have a bad afternoon.

A hand-drawn map titled who your store leans on. A box in the middle labeled your store has lines out to six circles: Shopify, payments, email and SMS, reviews, chat, and shipping app. A pencil note at the bottom says one status page each, one person on call.
Your list will differ. Write it down before the weekend, not during it.

You can’t prevent their outage. You can hear about it in minutes, instead of from a customer.

A phone showing a notification on a wooden bedside table in early morning light, a person still asleep in the bed behind it
The alert should wake one person, not the whole team.

Check this week: list every service in that chain and subscribe to each status page, starting with Shopify’s. Then decide who pauses the ads if checkout breaks.

Run yours on Monday, in 45 minutes

Our report is a starting point. Yours will be better, because your team already knows where your store is fragile.

  1. Book 45 minutes on Monday, November 16. Invite whoever runs ads, email and support, plus anyone who can touch the theme.
  2. Set the scene. “It’s Saturday morning. Black Friday went badly. What happened?”
  3. Write alone for five minutes. One failure per sticky note, each with a time of day.
  4. Read them out in turns, one note each, until nobody has any left.
  5. Give every note an owner and a check finished by Friday, November 20. A fix that needs new code waits for January, unless it repairs something broken.

Bring whoever answers support tickets. They already know where your store breaks.

Keep the list. On December 1, hold it up against what really happened.

The cheapest incident report is the one you write before the incident.

Want an outside read before the rush? Send us your store URL and within 24 hours we’ll send back a free report on your speed, checkout and conversion.