In the weeks before Black Friday, most teams run through creative calendars, email sequences and discount codes. Fewer run through the places a fast-moving store actually breaks under pressure: a payment method nobody tested with a live discount code, a tracking pixel that stops firing once a cart hits a certain size, an app update three agencies ago that nobody remembers approving. That worry is worth taking seriously, and the answer is simpler than it feels right now. Most Black Friday failures don't happen inside one system. They happen in the gap between two systems that were never tested together: your storefront and your payment processor, your apps and your tracking, your agency's calendar and your developer's.

The stakes are real and they're concentrated into a few hours. Adobe projects Cyber Week 2026 (Thanksgiving through Cyber Monday) will drive $47.5 billion in US online spend, up 7.4% year over year, with Black Friday alone at $12.9 billion and Cyber Monday at $15.1 billion. Last year, during the peak hours of 8pm to 10pm on Cyber Monday, US shoppers spent $16 million every minute. That's not one store's number. It's a sign of how little room there is to recover once the weekend starts. And mobile is where most of that spend happens: Adobe expects 63% of Thanksgiving 2026 online spend to come through a phone.

Here are six things worth checking before the weekend, in business terms, not developer terms.

How fast does your store load on a phone?

Speed is the first thing a shopper notices and the first thing most readiness checks skip, because the store "feels fine" on a laptop with good wifi in an office. Test it on an actual phone, on a slower connection, the way a real shopper will open it from a social ad or a text message during a sale.

The data on this is blunt. Sites that load in one second convert at an average of 3.05%, against 1.12% at three seconds and 0.67% at four. A one-second site converts about 2.5 times better than a five-second one, and conversion drops roughly 0.3 percentage points for every additional second of load time. On mobile specifically, a 0.1-second improvement in load speed was associated with an 8.4% increase in retail conversions and a 9.2% increase in average order value. Half a second of patience, multiplied across a weekend's worth of traffic, is a real amount of revenue.

We saw this firsthand building the live voting platform for The Game Awards, which had to hold up under 103 million views and over 45 million votes with almost no interruption. The lesson carries over directly: a store that performs well on a normal Tuesday and a store that performs well during a traffic spike are not tested the same way, and only one of those tests tells you what Black Friday will actually look like.

• Load your store on an actual phone, on a slower connection, not office wifi.
• Time it during a simulated rush, not a quiet moment.
• Measure against one second. Every extra second shows up in conversion.

Does checkout work on mobile with every discount code and payment option?

With most of this weekend's spend happening on phones, checkout has to work on a phone, with every payment method you offer and every discount code that's actually live for the sale. Not the general flow. The specific combinations: the buy-now-pay-later option plus a stacked discount, the saved card on a phone that auto-fills the wrong field, the promo code typed with a trailing space.

The average documented cart abandonment rate is 70.22%, and two of the most common named reasons are a website that errored or crashed (17% of shoppers) and a checkout that was too long or complicated (17%). The upside is just as concrete: better checkout design is associated with an average 35.26% potential increase in conversion for large ecommerce sites. That gap, between a checkout that technically works and one that's actually been tested the way shoppers will use it, is where a lot of Black Friday revenue goes missing this time of year.

• Run every live discount code through checkout on a phone, not just on a desktop.
• Test every payment and buy-now-pay-later option together, not one at a time.
• Check saved cards and autofill on mobile specifically.

What happens when something stops working?

Every store throws errors. The question that matters is what the shopper sees and does next. A clear, specific error ("that code expired," "try a different card") keeps someone in the flow. A blank screen or a generic failure sends them to a competitor's tab, and they usually don't come back.

This is also where small bugs turn into real cost at volume. We fixed a critical inventory overbooking bug for Something Borrowed Blooms, a flower subscription brand serving more than 40,000 customers, before it could turn into orders the business couldn't fulfill. A bug like that is invisible on a slow day and expensive on a fast one. Worth testing specifically: what happens when an item sells out mid-checkout, when two people grab the last one at once, and when a payment is declined partway through.

Fixing issues like this pays off beyond avoiding a bad moment. Cdiscount, a major French ecommerce retailer, improved its Core Web Vitals and saw a 6% revenue increase during Black Friday as a result.

• Trigger a few failures on purpose: a declined card, an expired code, a sold-out item mid-checkout.
• Read the actual error message a shopper would see. Vague means lost shoppers.
• Check what happens when two people buy the last unit at the same time.

What's frozen, and who can still change the site?

A code freeze is a period, usually starting a week or two before the sale, when nobody makes changes to the live site without a specific, approved reason. It exists because even a small, well-intentioned update (a new banner, a pricing tweak, an app upgrade) can break something else on the busiest day of the year, and nobody wants to discover that during the sale itself.

The business question underneath it is simple: who, right now, can log into your storefront, your apps and your payment settings and make a change? If the honest answer involves three agencies, a freelance developer and someone's personal login from eighteen months ago, that's worth fixing before the freeze starts, not during it. A freeze only works if everyone with access actually knows it's in effect and what counts as an emergency exception.

• List everyone who can currently log in to your storefront, apps and payment settings.
• Confirm the freeze dates and exception process, out loud, with every agency and vendor involved.
• Remove any access nobody can explain.

If something breaks, who do you call?

This is the short version of a longer question. If your storefront, your checkout app and your payment processor are run by three different vendors, and something breaks on Black Friday night, the honest answer to "who owns this" is often "we're not sure yet," and that uncertainty costs you the hour it takes to even figure out who to call. We'll cover what full ownership actually looks like, and how to set it up before next year, in the next post in this series. For now, the test is simple: write down who you'd call first, tonight, if checkout went down. If you can't answer in one sentence, that's the gap to close before the weekend.

• Write down who you'd call first, tonight, if checkout went down.
• Check it's one name, not "whoever's on call this week."

Can you trust your numbers this weekend?

The weekend you're about to run is also the one where you'll make next year's decisions, based on whatever your dashboards tell you. If your tracking doesn't match your actual orders and actual revenue, you'll walk away with the wrong lessons. Check now whether your analytics, your ad platforms and your order totals agree with each other on a normal day, so you have a baseline before the numbers get noisy.

It's also worth a quick look at where your traffic is actually coming from this year. Adobe recorded a 670% jump in US retail traffic arriving from AI shopping assistants on Cyber Monday last year, from a small starting base. A visit that starts with an AI assistant can look different in your analytics than a visit from a search ad, and if your tracking setup assumes every visit looks the same, you may be undercounting a channel that's growing fast.

None of this replaces having clean source data in the first place. Gartner estimates that poor data quality costs organizations at least $12.9 million a year on average. A dashboard built on bad inputs will tell you a confident, wrong story, and Black Friday weekend is an expensive time to learn that.

• Compare your dashboard totals to actual orders and revenue on a normal day, before the weekend gets noisy.
• Check how AI-assistant traffic shows up in your analytics, not just search and social.
• Confirm your source data is clean before you trust any report built on it.

A word on limits. This check is built for a store that's already live and handling real holiday traffic, not one mid-migration to a new platform or launching new checkout flows days before the sale. If that's your situation, the priority is different: freeze scope, not just code, and get a second set of eyes on anything that changed in the last month.

What to write down, during and after the weekend

The checklist above gets you ready. What you log during and after the weekend is what makes next year better than this one. Keep a running note, shared with whoever owns each piece, of three things:

1. What broke, and where it actually lived. Not "checkout was slow," but "checkout was slow specifically when the buy-now-pay-later option was selected on mobile."
2. Who found it, and how long it took to get to the right person. If the answer is "an hour, because we didn't know who owned it," that's next year's fix, not this year's.
3. What your numbers said versus what actually happened, so you know which dashboards to trust next time and which ones need a second look.

That list is the start of next year's roadmap. Most stores do this check once, under pressure, and then forget everything they learned the moment the sale ends. The ones that hold up next year are the ones that wrote it down this year.

If you want a second set of eyes on your stack before the weekend, contact us for a site audit.