- Before Black Friday, test the elements that sit on the path to the buy button: the banner, the discount display, the shipping message, trust badges, delivery hints and the mobile call to action.
- Rank every candidate by impact, traffic and time left. A test that cannot end by the end of calendar week 46 has no time to roll out before the peak.
- Black Friday 2026 is Friday, November 27 (week 48). Search interest peaks in weeks 46 and 47, so tests that finish earlier give you a full week to ship the winner.
- The page lists 12 tests with hypothesis, page and metric, a scoring grid and a calendar for weeks 43 to 48. Every runtime is a hypothesis. Check yours with the runtime calculator.
Before Black Friday, A/B test the elements that decide whether a visitor buys: how the offer is shown, the shipping threshold message, trust badges, delivery hints and the mobile buy button. Rank each test by impact, traffic and the time you have left, and stop all tests by the end of calendar week 46 so that the winners can be live in week 47.
Black Friday 2026 is Friday, November 27, and Cyber Monday follows on Monday, November 30. Below you find 12 tests with a hypothesis, a page and a metric each, a simple scoring grid, the Varify features that help, the mistakes that spoil pre-season tests and a calendar for weeks 43 to 48.
Why the weeks before Black Friday are your best testing window
Black Friday 2026 falls on Friday, November 27, in calendar week 48. Search interest for terms such as "black friday checklist", "black friday landing page" and "black friday strategy" rises from October and peaks in weeks 46 and 47 (Varify analysis of Google Trends, 2022–2024). So the visitors you test on in October and early November already include people who are planning their purchases.
Hypothesis: traffic grows in October and November, and a test that starts in week 43 or 44 collects data from this period. Check this in your own analytics before you rely on it. The practical benefit is time. A test that ends by week 46 gives you one full week to roll out the winner, check it on a phone and fix problems before the peak. A test that ends in week 48 produces a result you cannot use for this sale.
Use the test runtime calculator first. It tells you how many weeks a test needs for your traffic and conversion rate. If the answer is longer than the time you have left, choose a different test or a different page. The significance calculator helps you read the result when the test ends. For the basics, see the A/B testing guide.
12 tests to run before Black Friday
The list below mixes quick wins and tests that need more time. Each row has a hypothesis, the page it runs on and the metric that decides it. All runtimes are labeled as hypotheses because they depend on your traffic. Replace them with the output of the runtime calculator.
Your shop may need different tests than this list. Enter the URL of a page into the AI CRO Audit. It analyzes the page and returns CRO insights, concrete recommendations and 10 test ideas that work directly as A/B tests.
| # | Test | Hypothesis | Page | Metric | Minimum runtime |
|---|---|---|---|---|---|
| 1 | Banner headline | A banner that names the end date earns more clicks than one that names only the discount. | Site-wide banner | Clicks on the banner | Hypothesis: 1 week |
| 2 | Discount display | Showing the saving as an amount makes the offer easier to grasp than showing it as a percentage, or the reverse. | Category and product tiles | Add-to-cart rate | Hypothesis: 2 weeks |
| 3 | Free-shipping message | A shipping threshold line in the header and the cart lifts the order value. | Header, cart | Average order value | Hypothesis: 2 weeks |
| 4 | Trust badges | Payment and returns badges next to the buy button reduce hesitation. | Product page | Add-to-cart rate | Hypothesis: 2 weeks |
| 5 | Delivery date | A delivery date or last order date for Christmas moves more visitors into checkout. | Product page, cart | Checkout starts | Hypothesis: 2 weeks |
| 6 | Total price in the cart | Showing shipping and taxes in the cart removes surprises at the last step. | Cart | Checkout starts | Hypothesis: 2 weeks |
| 7 | Guest checkout hint | A clear line that no account is needed gets more carts through. | Cart | Purchases | Hypothesis: 2 weeks |
| 8 | Mobile call to action | A sticky buy button helps phone visitors act from any scroll position. | Product page, mobile | Add-to-cart rate | Hypothesis: 1 week |
| 9 | Sale badges | A badge on sale items leads more visitors to open them. | Category page | Clicks on products | Hypothesis: 1 week |
| 10 | Early-access teaser | A teaser with the start date brings visitors back on day one. | Home page | Clicks on the teaser | Hypothesis: 1 week |
| 11 | Bundle or gift module | A bundle or a free gift above a set value raises the basket. | Product page, cart | Average order value | Hypothesis: 2 weeks |
| 12 | Low-stock wording | A stock label that reflects real inventory speeds up decisions. | Product page | Add-to-cart rate | Hypothesis: 1 week |
Source: Varify working hypotheses. Runtimes depend on your traffic and conversion rate. Calculate yours with the test runtime calculator before you start.
Tests 5 to 7 sit close to the checkout and reward careful quality assurance. Use real data in labels such as stock levels or delivery dates, and have wording around deadlines, discounts and countdowns checked legally before the start. Tests 4, 5, 8 and 12 run on the product page. For more changes there, see Black Friday product page optimization.
How to prioritize: impact, traffic, time left
You will not run all twelve tests. Score each candidate from 1 to 3 on three questions and multiply the scores. The product ranges from 1 to 27. Start with the highest.
| Score | Impact | Traffic | Time left |
|---|---|---|---|
| 3 | The element sits on the direct path to the buy button. | The page gets most of your sale traffic. | The test can end by the start of week 46. |
| 2 | The element supports the decision but sits beside the path. | The page gets a fair share of sale traffic. | The test needs until the middle of week 46. |
| 1 | The element is easy to overlook. | The page gets little traffic. | The test would run into week 47 or later. |
Source: Varify scoring method. Multiply the three scores. Run the highest products first. The scores are a planning aid. They are not a statistical result.
Example scoring. The free-shipping message in the header and cart (test 3) scores 3 for impact, 3 for traffic and 3 for time left, a product of 27. A badge in the footer scores 1 for impact, 2 for traffic and 3 for time left, a product of 6. The first test goes into the calendar, the second waits. Use your own numbers. Impact is a judgment about the visitor's path. Traffic comes from your analytics. Time left comes from the runtime calculator.
A tie goes to the test with the shorter runtime. If two tests share a page, run them one after the other.
Running pre-Black-Friday tests with Varify
Varify gives you the pieces for a short test season in one place:
- Prompt-Based Experimentation (PBX). Describe the test in plain language, for example "show the end date in the banner headline". In the Visual Editor, "Ask AI" builds the variation, writes the code and sets up the experiment, without a developer. You can refine a variation as often as you like, so the first variations are ready in week 43 and every test can start on time. See the PBX examples and the PBX docs.
- Visual Editor and Code Editor. Fine-tune banners, badges, labels and layouts by hand. Developers can adjust the JavaScript code the AI generated. For ready-made banner elements, see how to test Black Friday banners without a developer.
- Preview link and QA mode. Show every variation to your team and confirm that goals fire. The quality assurance guide lists the steps.
- Schedule Experiments. Set a start and end date so that a test begins and ends without anyone watching the clock. See scheduling experiments.
- Bayesian statistics with remaining runtime. Read how much time a running test still needs.
- Personalization. Serve the winning variation to 100% of a chosen audience while your team updates the shop's own templates.
- GA4 integration and Shopify goals. Measure revenue, purchases and order value per variation.
Varify works without cookies, which keeps the setup for a short sale-season test simple. Your own legal and data protection review of your setup stays with you.
Mistakes that ruin pre-season tests
- Too many tests on the same page. Two variations that change the same area confuse the result. Run them one after the other, or use mutual exclusion so that a visitor takes part in only one of them.
- Ending a test by gut feeling. Decide the runtime and the metric before the start. Stopping the moment one variation looks ahead produces false winners.
- Changing the offer during a test. A new discount level or a new end date in the middle of a run mixes two effects.
- No rollout of the winner. A winning variation that stays a test variation disappears when the test stops. Plan the rollout in week 47.
- Testing only on desktop. Open every variation on a phone before the start. Many sale visitors use one (see mobile A/B testing).
- Starting too late. A test that cannot end by week 46 belongs in January.
Your test calendar until November 27
Plan backwards from Friday, November 27. The calendar below fits a shop that starts testing this week. Move the start dates if you are later, and drop the longest tests first.
| Calendar week | Dates in 2026 | What to do | Tests |
|---|---|---|---|
| Week 43 | October 19–25 | Check goals and tracking. Pick tests with the grid. Build the first variations and open the preview link. Start the one-week tests. | Tests 1, 8, 9, 12 |
| Week 44 | October 26–November 1 | Read the first results. Start the two-week tests on product pages and the cart. | Tests 2 to 7, 11 |
| Week 45 | November 2–8 | Roll out clear winners. Start the early-access teaser. | Test 10 |
| Week 46 | November 9–15 | Last reading of results. All tests end by the end of this week. | Evaluate |
| Week 47 | November 16–22 | Roll out all winners to all visitors. Freeze the shop. Check the end dates everywhere. | Ship |
| Week 48 | November 23–29 | Black Friday is Friday, November 27. Watch, and ship fixes only. | No new tests |
Source: 2026 calendar. Black Friday is November 27 (week 48), Cyber Monday is November 30 (week 49).
Also look at how a test behaves during the peak itself. The guide to A/B testing during Black Friday covers when running a test in the peak week is safe. Once the tests are chosen, check the page they run on with the Black Friday landing page checklist.
Run your pre-Black-Friday tests with Varify.
Prompt-Based Experimentation, preview link, QA mode and scheduled start and end dates. Plan the runtime with the free calculator first.