Thanksgiving falls on November 26 this year, which puts Black Friday on November 27 and starts the busiest few weeks of the year for many businesses. For retailers it is obvious. For everyone else — restaurants taking catering orders, service companies booking end-of-year appointments, event venues, nonprofits running year-end giving — it is often a surprise when the website that has been fine all year slows down on the one day it matters.
The good news is that most holiday outages come from a short list of causes, and most of them can be dealt with in a few weeks of focused work. Here is the checklist we use.
Know what "busy" means for you
Start with your own numbers. Look at last year's traffic, orders, or bookings for the same weeks and find your highest hour. Then decide what you are planning for this year: the same, double, or something else because of a promotion or a new marketing push.
That one number shapes everything else. A site that needs to handle twice an ordinary afternoon is a very different problem from a site that needs to handle twenty times one.
Find the bottleneck before your customers do
Every system has one piece that gives out first. In the small-business apps we work on, it is usually one of these:
- The database. A report or search page that runs a heavy query on every visit is fine at ten users and painful at two hundred.
- A third-party service. Your payment processor, shipping calculator, inventory system, or email provider may have its own rate limits. If your checkout waits on all of them, it is only as fast as the slowest.
- Images and media. Large, unoptimized images make every page heavier and can push you into higher hosting or bandwidth charges.
- A single small server. Hosting that was sized for a quiet month, with no way to add capacity quickly.
A load test — simulated traffic aimed at a staging copy of your site — is the most direct way to find out which one you have. It does not need to be elaborate. Even a simple test of your three most important pages and your checkout or booking flow will tell you a lot.
Cheap wins that usually help
Cache what does not change. Product pages, menus, and service descriptions do not need to be rebuilt for every visitor. A content delivery network or page caching can take a large share of the load off your server.
Compress and resize images. This is often the fastest improvement available, and it helps every visitor, not just the holiday ones.
Turn off what you do not need. Old tracking scripts, plugins nobody uses, and chat widgets from a trial that ended all slow pages down.
Check your hosting plan's limits. Know whether your plan can scale up automatically, how quickly, and what it costs. A surprise bill in January is better than an outage in November, but neither should be a surprise.
Protect the paths that make money
Not every page is equally important. A slow blog post is an annoyance. A broken checkout or booking form is lost revenue.
Identify the two or three flows that matter most and make sure they keep working even when something else does not. That might mean a fallback if a shipping-rate service times out, a queue for confirmation emails so a slow email provider does not hold up the order, or a simple "we're busy, try again in a minute" page instead of an error.
Freeze changes before the rush
Most outages we see during busy periods are caused by a change, not by traffic. Someone pushes a small update on a Friday afternoon and something unrelated breaks.
Set a change freeze. A common pattern is to finish all planned work by mid-November and allow only urgent fixes through the end of the year. Promotions, banners, and price changes that need to go live during the freeze should be built and tested in advance and switched on, not deployed.
Decide who is on call
Before the busy weeks, answer these questions in writing:
- Who gets alerted if the site goes down or checkout errors spike?
- How do they get alerted — text, phone call, or an email nobody reads on a holiday?
- Who has the access needed to fix things, including hosting, domain, and payment accounts?
- What is the phone number for each vendor's support line?
- Who decides whether to roll back a change?
If the answer to any of these is "the developer we used last year," find out now whether that person is still reachable.
Monitor what customers experience
Server dashboards are useful, but the most important signal is whether a customer can actually complete a purchase or booking. An uptime monitor that loads your checkout page every few minutes, plus an alert when orders drop to zero during business hours, will catch problems that server metrics miss.
Watch for fraud, too
Busy periods attract card testing — automated attempts to run stolen card numbers through small purchases — and fake orders. Make sure your payment processor's fraud tools are turned on and that someone looks at unusual order patterns.
A two-week plan
If you are starting now:
- Week one: pull last year's peak numbers, list your critical flows, run a basic load test, and fix the worst bottleneck.
- Week two: optimize images, enable caching, confirm hosting limits, set up monitoring and alerts, write down the on-call plan, and start the change freeze.
We have spent a lot of time on systems that need to hold up under load, from consumer apps to the Xsatori exchange. If you want a second set of eyes on your site or app before the rush, tell us what you're running and we will help you find the weak spot first.