Industry Guide Technology
Four weeks before Black Friday, the work is testing and ownership, not new features. A checklist for checkout testing, dependencies, alerting and incident response.
In 2026, Thanksgiving falls on November 26, Black Friday on November 27 and Cyber Monday on November 30. With roughly a month to go, the most valuable infrastructure work is no longer building capacity from scratch. It is proving that the checkout works under realistic load, knowing which outside services can break it, and making sure a named person responds when something does.
Use the checklist below in order. Each section ends with the question that tells you whether it is done.
| Window (2026) | Focus |
|---|---|
| Now to November 6 | Dependency map, load and spike tests, quota increases |
| November 9 to 13 | Fix what testing found; retest; agree alert owners and rota |
| November 16 to 20 | Change freeze begins; runbook walk-throughs; rollback test |
| November 23 to December 1 | Peak operations: monitoring, daily check-ins, approved changes only |
| December 2 to 4 | Post-event review while details are fresh |
Adjust the dates to your own release calendar. The principle is to finish testing early enough to fix what it finds, and to freeze changes before traffic builds rather than on the day.
A fast homepage tells you little about Black Friday website performance. Revenue depends on a chain: product page, search, cart, shipping calculation, tax, payment authorization, order confirmation and the email that follows.
Traffic does not arrive only on Friday. LogicMonitor's peak-season readiness guidance, published October 16, 2026, cites Fastly analysis showing traffic rising before Black Friday and staying elevated through the month, which is why tests should cover sustained load, not just a one-hour peak.
Done when: you know the order rate at which checkout times or error rates start to degrade, and it is comfortably above your forecast.
Most checkouts depend on services you do not run. List every one a purchase touches: DNS, CDN, web application firewall, hosting or platform provider, search, payments, fraud screening, tax, shipping rates, inventory sync, analytics and tag manager scripts.
| Dependency | Question to answer | Fallback if it fails |
|---|---|---|
| Payment gateway | Is a second method or provider available? | Hide the failing method; show a clear message |
| Shipping or tax API | What happens if it times out? | Flat-rate shipping or a cached rate table |
| Site search | Can browsing still work if search is slow? | Category navigation and popular-product links |
| Third-party scripts | Do any block page rendering? | Load asynchronously or disable through tag manager |
| CDN or DNS | Who can switch configuration, and how fast? | Documented change with an owner and approval |
Outages at shared providers are a real risk. The same LogicMonitor article notes that a Cloudflare network failure began on November 18, 2025, ten days before last year's Black Friday, and affected retailers relying on its services.
Check capacity limits as well as servers: database connection limits, API rate limits with partners, autoscaling maximums, and account-level quotas on your cloud provider. Raise quotas now; requests can take time to approve.
Done when: every dependency on the checkout path has an owner, a known failure behavior and an agreed fallback.
Alerts without owners create noise, not responses.
Done when: a payment-failure spike at 2 a.m. on Saturday would reach a named person within minutes.
Decide in advance what you will do, so nobody improvises under pressure.
Done when: each person on the rota knows which runbook they own and has walked through it at least once.
After the peak, compare actual results with your baseline: checkout success rate, payment authorization rate, page response time at the busiest hour, the number of incidents, and the time from alert to response. Note which dependencies caused problems and which fallbacks you used. Those notes become next year's checklist and often reveal architecture work worth scheduling for the quieter first quarter.
If campaigns are part of your peak plan, align send times with this checklist; a large email drop is effectively a scheduled traffic spike. Our Black Friday email and SMS checklist covers the marketing side.
In 2026, Thanksgiving falls on Thursday, November 26, Black Friday on November 27 and Cyber Monday on November 30. Traffic typically builds before Black Friday and stays elevated through the weekend, so plan monitoring and on-call cover for the whole window, not just one day.
Script the three to five journeys that produce most revenue, including guest checkout and a discount code, and test the full path from product page to order confirmation. Run sustained load above your forecast peak plus a short spike test, use your payment provider's test mode, and warn third parties before sending test traffic to them.
One page per likely failure, such as a degraded payment provider, slow search, delayed inventory sync or CDN problems. Each page should name the owner, the first checks to run, the fallback, who can switch off non-essential features and how to communicate with customers and the wider team.
Growix's Infrastructure Reliability & Monitoring service sets up monitoring, alerting and incident troubleshooting, and our Cloud Infrastructure & Architecture team reviews scaling and capacity. With weeks rather than months remaining, the realistic scope is testing, alerting and runbooks, not re-architecture.
Token prices are only part of the bill. Model the full monthly cost of an AI feature, including retries, monitoring and human review, and measure cost per successful task.
Lower the cost of serving each customer, not just the monthly bill. A four-step method for SaaS teams on AWS, Azure or Google Cloud, with a worked example.
Migration fixes platform limits; optimization fixes waste and performance. Diagnose which constraint you have, then sequence the work so you never migrate what you should retire.
