Growth / Seasonal
BFCM Technical Readiness Checklist
// peak traffic finds every crack you didn't fix
By Fracto Solutions
October 30, 2026
The short answer
Before BFCM, confirm the technical basics that peak traffic will test: site speed and stability under load, a flawless checkout and payment flow, accurate tracking and analytics, apps and theme that won’t buckle, and a change freeze before the weekend. A fast, stable, well-tracked store is what turns record traffic into record sales instead of record abandonment.
Why does technical readiness matter at peak?
Because the same defect costs far more when the traffic is concentrated. That much is arithmetic. What is not established, and what we used to imply, is that shoppers become more sensitive to slowness under load, so we have stopped asserting it.
The honest version is simpler and still compelling. Shopify merchants peaked at $5.1m in sales per minute at 12:01pm EST on Black Friday 2025, and Adobe measured $16m per minute across US retail during the Cyber Monday evening peak. A checkout fault during those windows costs orders at that rate, whatever the shopper’s psychology. Documented average cart abandonment is already 70.22%, averaged across 50 studies from 2006 to 2025, before peak load enters the picture.
So the case for readiness work is about exposure, not about a special peak-traffic effect. Fix it in early November when a mistake is cheap, rather than on the Friday when it is not.
What should I check, and against what numbers?
Use published thresholds rather than adjectives. A checklist without a pass mark cannot be failed, which is why so many stores complete one and still go into the weekend slow.
Check
Pass mark
Source of the threshold
Largest Contentful Paint
2.5s or less, 75th percentile
Google Core Web Vitals
Interaction to Next Paint
200ms or less, 75th percentile
Google; replaced FID on Mar 12, 2024
Cumulative Layout Shift
0.1 or less, 75th percentile
Google Core Web Vitals
Load test
At least 2x your own prior peak orders per minute
Your analytics, not a generic number
Checkout
End to end on real mid-range mobile, every payment method
Practitioner standard
Change freeze
On or before Friday Nov 20, 2026
Working back from Nov 26 peak
Assess the vitals on real-user mobile data at the 75th percentile, not on a lab run from your own laptop.
Beyond the vitals, audit app weight. Every app that injects script into the storefront costs something on every page view, and the cost compounds on mobile. List what is installed, identify what each one adds to the critical path, and remove anything you cannot justify before the freeze. Then verify tracking: analytics, conversion tags and server-side events, tested with a real transaction rather than a preview tool, because a weekend measured wrong is a weekend you cannot learn from.
How do I test under realistic conditions?
On the devices and connections your shoppers actually use, at a concurrency sized from your own history. Testing on office wifi with a desktop browser is how stores conclude they are fast.
Mobile is the majority and still growing: Adobe put mobile at 56.4% of 2025 season transactions, up from 54.5% the year before, and 57.5% of Cyber Monday sales; Salesforce measured 70% of online orders placed on mobile globally. So test on a mid-range Android handset over throttled mobile data, and treat that result as the real one.
For load, size the test from your own peak rather than a round number. Take last year’s highest orders-per-minute and highest concurrent sessions, double them, and confirm the store holds. If you have never measured it, this year’s instrumentation is the deliverable; capture it so next year’s test has a basis.
On what to expect from the speed work, keep the evidence in proportion. Deloitte’s study for Google found a 0.1 second mobile improvement associated with retail conversion up 8.4% and spend up 9.2%, across 37 large brand sites and over 30 million sessions in 2019, but it was observational rather than randomised and required improvement across all four of its measured metrics. Akamai’s 2017 analysis of roughly 10 billion visits put a 100ms delay at “as much as 7%” of conversion, an upper bound that gets quoted as a point estimate. Useful direction; not a formula to forecast with. And the widely repeated “Amazon found 100ms costs 1% of sales” has no published study behind it at all, so we do not cite it.
What should a change freeze cover?
Everything that can alter the storefront or the checkout: deploys, theme edits, app installs and removals, payment configuration, and script changes. Pick the date, write it down, and make it someone’s job to enforce.
For 2026, freeze on or before Friday November 20. That leaves a working week to catch anything the freeze itself surfaces, and it is far enough from the November 26 peak that a late problem is still fixable in daylight.
Write down three things alongside the date. What is exempt, which should be a very short list, typically content and price changes through an established, tested path. Who can authorise an exception, which should be one named person. And what the rollback is for each exempt change, prepared before the change goes out rather than improvised after it.
The freeze is also when you set up monitoring you will actually watch: uptime and checkout success rate, orders per minute against your load-test ceiling, error rates, and payment gateway failures. Decide now what number triggers a phone call, because at 2am on Black Friday nobody wants to be deciding that for the first time.
We pressure-test the store before the weekend, then lock it down.
We test speed and checkout under realistic load, confirm tracking is firing, audit apps and theme for weak points, verify inventory sync, and set a change freeze, so the store goes into BFCM proven and stable rather than hopeful.
It’s the same measurement-first approach behind our web development and conversion optimization work, where we rebuild the buying path and measure the result .
// we would rather show up in your reporting than in your inbox
Frequently asked questions
How fast does my store need to be for BFCM?
Use Google’s published Core Web Vitals pass marks rather than a target that sounds fast: Largest Contentful Paint at or under 2.5 seconds, Interaction to Next Paint at or under 200 milliseconds, and Cumulative Layout Shift at or under 0.1, all at the 75th percentile of real-user traffic and split by mobile and desktop. If your checklist still measures First Input Delay, INP replaced it on March 12, 2024.
Does site speed really affect conversion?
The evidence points that way but is narrower than its reputation. Deloitte’s study for Google associated a 0.1 second mobile improvement with retail conversion up 8.4%, across 37 large brand sites and 30 million sessions, and it was observational rather than a controlled experiment. Akamai put a 100ms delay at up to 7% of conversion. Neither tested a peak event, so treat them as direction rather than a formula.
How should I size a BFCM load test?
From your own history, not a generic figure. Take last year’s peak orders per minute and peak concurrent sessions, double them, and confirm the store holds. If you have never captured those numbers, instrument this year so next year’s test has a baseline.
When should I freeze changes before BFCM?
On or before Friday November 20, 2026, which leaves a working week before the November 26 peak. Cover deploys, theme edits, app installs and removals, payment configuration and script changes. Name one person who can authorise an exception, keep the exempt list very short, and prepare the rollback before any exempt change ships.
Sources
Google/Chrome, “Web Vitals” (web.dev), last updated October 31, 2024. The “good” thresholds, assessed at the 75th percentile of page loads and segmented by mobile and desktop: Largest Contentful Paint at or under 2.5 seconds, Interaction to Next Paint at or under 200 milliseconds, Cumulative Layout Shift at or under 0.1. Google Core Web Vitals guide
Google/Chrome, “Interaction to Next Paint becomes a Core Web Vital” (web.dev). INP officially replaced First Input Delay as a Core Web Vital on March 12, 2024. Any readiness checklist still targeting FID is two years out of date. Google INP Core Web Vitals announcement
Deloitte for Google, “Milliseconds Make Millions” (page updated June 24, 2020; data collected end of 2019). 37 leading European and American brand sites, more than 30m user sessions, mobile load times monitored hourly for 30 days. A 0.1 second improvement in mobile site speed was associated with retail conversion up 8.4% and retail spend up 9.2%, and travel conversion up 10.1%. This is observational, not a controlled experiment: there was no random assignment, the improvement had to occur across all four measured metrics, one of which (First Meaningful Paint) is now deprecated, and the sample is 37 large brands. web.dev case study: milliseconds make millions
Akamai/SOASTA, 2017 analysis of roughly 10bn user visits over one month across top online retailers. A 100ms delay was associated with a hit to conversion of “as much as 7%,” and a 2 second delay with a 103% increase in bounce rate. Correlational, an upper bound rather than a point estimate, the site set is undisclosed, and it is now nine years old. Retail TouchPoints on speed and conversions
Baymard Institute, cart abandonment rate, updated September 22, 2025. A 70.22% average documented cart abandonment rate. This is an average across 50 separate studies published between 2006 and 2025, not a single measurement, and should be described that way. Baymard Institute cart abandonment rate statistics
Shopify, BFCM 2025 results, published December 2, 2025. $14.6bn global GMV, up 27% year on year (24% constant currency); average cart price $114.70; cross-border orders 16% of all global orders; peak of $5.1m in sales per minute at 12:01pm EST on Black Friday; 94,900+ merchants recorded their highest-selling day ever; 81m+ customers, counted by unique buyer email; top countries US, UK, Australia, Germany, Canada. Platform-measured but self-reported, and Shopify states the data is unaudited and approximate. Shopify does not publish a BFCM mobile share, so any such figure attributed to it is not real. Shopify BFCM 2025 data
Adobe Analytics, 2025 US holiday season final report, published January 7, 2026. Measured transaction data across more than 1tn visits to US retail sites, 100m SKUs and 18 product categories. Cyber Week (Nov 27 to Dec 1, 2025) $44.2bn, up 7.7%; full season Nov 1 to Dec 31 $257.8bn, up 6.8%; mobile 56.4% of season transactions, against 54.5% in 2024; buy now pay later $20bn for the season, up 9.8%; peak discount depths by category, from electronics at 30.9% down to furniture at 18.8%; and 25 days of the season exceeded $4bn, up from 18 days in 2024. Adobe 2025 holiday shopping report
Adobe Analytics, Cyber Monday 2025 release, published December 2, 2025. Cyber Monday $14.25bn, up 7.1%, the largest online shopping day on record, peaking at $16m per minute between 8pm and 10pm; Black Friday $11.8bn, up 9.1%; Thanksgiving $6.4bn, up 5.3%; mobile 57.5% of Cyber Monday sales; buy now pay later $1.03bn on Cyber Monday alone. Same measurement basis as above. Adobe Cyber Monday record data
Salesforce, Cyber Week 2025 results, published December 5, 2025. Aggregate platform data across more than 1.5bn global shoppers in 89+ countries, for November 25 to December 1, 2025. Global Cyber Week $336.6bn, up 7%; US $79.6bn, up 5%; mobile drove 70% of online orders both globally and in the US; AI and agents drove $67bn in sales and influenced 20% of global orders. Salesforce publishes average selling price growth rather than AOV, and publishes no Cyber Week conversion rate or average discount rate, so we do not quote those. Salesforce Cyber Week 2025 data Shopify, “Black Friday Checklist 2026: 25 BFCM Prep Steps,” updated August 28, 2026. Vendor guidance. Cited for the third-party figures it carries and attributes: Baymard’s 39% of abandoners citing unexpected extra costs; Narvar 2025, 55% of consumers saying on-time delivery makes them more likely to buy again; and its own concrete timing recommendation that thank-you emails be sent 5 to 7 days after the event. Shopify BFCM checklist