U.S. holiday eCommerce sales are on track to hit $316.1 billion to $318.9 billion between November and January, according to Deloitte.
That’s a growth of 7.5% to 8.4% that outpaces the 4.0% to 4.8% forecast for total holiday retail.
That difference shows that shoppers are increasingly doing their holiday spending online, and they’re not loyal while they do it.
They’re comparing prices across tabs, switching retailers mid-cart and chasing the better deal.
A slow product page or a broken discount code doesn’t just cost a sale. It hands that sale to whoever loads faster or checks out cleaner.
That reality puts brands and their agencies in an uncomfortable spot every October. There’s a long list of things worth fixing on any eCommerce site at any time of year.
The question that actually matters right now isn’t what needs fixing. It’s what’s safe to touch three weeks out from Black Friday, and what should sit untouched until January.
When Should a Site Stop Taking Updates Before Black Friday?
A change that touches checkout, payment processing, or inventory sync has a blast radius that includes every single transaction on the site.
A change to a blog layout or an About page has a blast radius of basically nothing. The size of that radius, not how good the idea is, decides if it ships this month.
Most teams get this backward. They rush the flashy stuff, a new homepage hero or a redesigned nav, because it’s visible and somebody on the marketing team wants it live before the big weekend.
Meanwhile, the things that drive revenue during peak traffic, like checkout speed and mobile load times, get treated as routine maintenance and pushed to whenever there’s time.
I’d flip that order every time.
We saw what that order looks like done right when we built McLaren’s new golf storefront on Shopify earlier this year.
The brand was carrying a premium racing identity into a retail category where buyers had no existing relationship with it. We focused on:
- Checkout speed. Keeping the path from browsing gear to purchasing short and frictionless.
- User journey. Prioritizing speed and clarity before adding brand storytelling.
- Brand experience. Carrying McLaren’s racing identity into the storefront without distracting from the purchase.
Sales velocity came in ahead of initial projections once the site went live.
That’s the order I’d protect going into any holiday season. The freeze date for anything touching checkout or payments should land two to three weeks before your first major traffic spike.
After that date, changes should be limited to fixing broken functionality, not improving things that already work.
Which Fixes Are Worth the Risk This Close to Peak?
Mobile is where I’d spend whatever time is left. Adobe expects mobile devices to account for 57.4% of U.S. online holiday spending in November and December, rising to 63% on Thanksgiving itself.
That means more than half the traffic hitting a site during its biggest weekend of the year is coming through a phone, often on a spotty connection in a crowded store or at a dinner table.
If a product page takes four seconds to load on mobile, or a checkout form makes someone pinch and zoom to find the submit button, that’s the fix worth making now.
It’s low risk, because you’re tightening something that already exists rather than introducing something new, and it directly protects revenue on the channel carrying the most weight.
Lounge Lizard's OpenWorks website launch also demonstrates how the agency approaches website redesigns with a focus on communicating a company's value proposition, culture, and customer experience:
Promotions deserve the same scrutiny, for a different reason.
Deloitte’s 2026 Holiday Retail Survey found that 38% of U.S. consumers qualify as value seekers, meaning they use five or more money-saving behaviors during the season, and these shoppers expect to spend 12% more than everyone else.
That’s not a small, price-sensitive niche. That’s over a third of the holiday shopping audience actively hunting for codes, rewards, and private-label alternatives, and planning to spend more because they found them.
If a discount code fails at checkout for that shopper, you haven’t just lost a sale.
You’ve lost a shopper who was ready to spend above average and walked because the system didn’t honor what it promised.
Before Black Friday traffic arrives, I’d run every active promotion, every stacking rule, and every third-party app connected to checkout through a real stress test, not a quiet afternoon click-through.
Simulate concurrent users. Try to break the discount logic on purpose. Find out now what a support ticket flood would tell you in December.
Brands that migrated to a new platform this year carry an extra variable.
Their first peak season is also their first real load test, and load behavior under five times normal traffic rarely matches what a staging environment showed.
I’d want those teams watching server response times and third-party app performance in real time through the season, not just reviewing dashboards after the fact, because a platform that felt fast in September can buckle in a way nobody predicted once Black Friday traffic actually lands.
None of this should be treated as a one-time cleanup before the season starts.
The sites that come out of January ahead of where they started are the ones tracking what actually happened during peak:
- Where shoppers abandoned carts
- Which promotions drove orders versus which ones just drove clicks
- Where mobile checkout friction showed up in the data instead of in a guess
That record is what should drive the list of priorities for February, not whatever felt urgent back in October. The freeze protects the season.
What you measure during it is what earns you a better one next year.
A site that holds steady through December and comes out the other side with a clear read on what shoppers actually did will outperform one that spent the season guessing and the new year catching up.
