The short answer
- On Shopify Plus, checkout itself is testable. Trust badges, a delivery promise or a cut-off notice inside checkout can be tested in October and kept through the peak if they win.
- On every plan except Starter, the thank-you page is testable. A code for the next order, shown after payment, is the cheapest way to bring Black Friday buyers back in December.
- Post-purchase upsells (one click, after payment, before the thank-you page) need Shopify's approval for each app on live stores. Measure the take rate and the refunds, not just the extra revenue.
- Checkout changes are freeze items. Whatever you test here must end, and be shipped or dropped, before Monday 16 November (part 3).
Who this is for: Shopify brands doing roughly $1M to $10M a year, on Plus or on Basic, Shopify or Advanced, deciding what to change in and after checkout before the sale.
The 2026 timeline
| When | What happens |
|---|---|
| Tue 27 Oct to Fri 30 Oct | Pick the checkout or thank-you test. Place the app blocks in the checkout editor. QA on a phone. |
| By Mon 2 Nov | Start. Checkout tests on Plus reach their sample quickly, but still need 7 full days. |
| Thu 12 Nov | Tests end. Read the result. |
| Fri 13 Nov | Keep the winning blocks in place, or remove the losing ones. |
| Mon 16 Nov | Freeze. No checkout block changes after this. |
| December to February | Read thank-you page codes: repeat orders at 30 and 90 days. |
What Shopify lets you change, by plan
| Where | Plans (Shopify developer docs, read 12 October 2026) | What it's good for |
|---|---|---|
| Inside checkout (information, shipping and payment steps) | Shopify Plus only | Reassurance and clarity at the moment of paying |
| Thank-you and order status pages | Every plan except Shopify Starter | The next order: codes, referrals, a one-question survey |
| Post-purchase page (between payment and thank-you) | Every plan except Starter; in beta, and each app needs Shopify's approval for live stores | One-click upsells into the order just placed |
Prices never change in checkout tests. Testing prices, offers and shipping is covered in parts 4 to 6.
Step 1 (Plus): Test reassurance in checkout
A shopper who reached checkout has already decided they want the product. What stops them now is doubt: about delivery, returns, or whether to trust the store with a card. Three ideas fit Black Friday, each with its mechanism:
- A delivery promise. "Order by 15 December for delivery before Christmas" answers the question that drives late-November orders. It works only if you can keep it, in every market you ship to.
- Trust badges or a returns line. Useful for first-time buyers from paid traffic, who don't know your brand. Returning customers ignore them, which dilutes the effect.
- A cut-off notice. For a sale that ends at a set time, a short line in checkout ("Black Friday prices end Monday 30 November at midnight ET") must match the banner and the discount's real end date exactly.
What to measure. The primary goal is purchases per shopper who reached checkout, or revenue per visitor. Step completion (contact, then shipping, then payment) is a diagnostic: it tells you where a version helped, never whether it won. Payment errors and support contacts are guardrails.
Shoppers who skip the cart. Buy it now and accelerated checkout buttons on a product page can bypass the step that assigns a version. Check how your tool treats those shoppers, and that the rule is the same for every version.
Illustrative inputs, not data from a real store. Computed with the same sample-size formula Benchmyrk's test planner uses (two-sided, 95% confidence, 80% power, two versions, even split).
Inputs: 1,200 shoppers a day reach checkout (out of about 8,571 store visitors a day, as in part 1), and 55% of them complete a purchase.
| Smallest change in checkout completion to detect | Shoppers per version | Days |
|---|---|---|
| 10% (55% to 60.5%) | 1,265 | 3, but no verdict before 7 full days |
| 5% (55% to 57.75%) | 5,104 | 9 |
Checkout tests on a brand's traffic are fast, because completion rates are high and every shopper in the test is close to buying. That makes Plus checkout one of the few places where a two-week test before the freeze can detect a modest change. Seven full days is still the minimum, because weekday and weekend shoppers differ. The sample size and test duration guide explains the inputs.
Step 2 (every plan except Starter): Test the thank-you page for December
Black Friday brings you more first-time buyers than any other week. Most of them came for a deal, and many won't come back. The thank-you page is the one moment you have their full attention after they paid.
Test what you show there:
- A code for the next order with a short expiry (for example, valid in December). It gives a reason to come back before the holidays.
- A referral link. Works best for products people talk about, less well for replenishment items.
- A one-question survey ("How did you hear about us?"). No sales effect, but it fills a gap in attribution during the busiest week of the year.
What to measure. The primary goal is repeat purchases within a set window, chosen before launch. Clicks and code copies are diagnostics. Refunds and cancellations of the later orders are guardrails. The order that showed the thank-you page should never count, since it was placed before anyone saw the test.
Plan the read for January. A thank-you test started in November can't be read before the window closes. With a 30-day window, buyers from Cyber Monday are counted until the end of December.
Illustrative inputs. Computed with the same formula as above, with a 10% baseline rate of buyers who order again within the window.
| Buyers per version | Smallest change in repeat-purchase rate the test can detect |
|---|---|
| 2,000 | 28.2% (10% to about 12.8%) |
| 3,000 | 22.8% |
| 5,000 | 17.5% |
A brand with 6,000 orders over the BFCM weekend, split evenly, has about 3,000 buyers per version. That detects only a large change. A small code that wins by 10% will most likely read as "inconclusive". Decide before launch what you'll do then (keep the simpler version, or the one with the lower discount cost).
Step 3: Post-purchase upsells, with care
A post-purchase upsell shows one product after payment. One click adds it to the order the buyer just placed, charged to the same payment method. Two things to know before you plan on it:
- Approval. Shopify's post-purchase page is in beta, and each app needs Shopify's approval to use it on live stores. Check that your upsell app has it before you plan the sale around it.
- Refunds. An upsell accepted on impulse is returned more often than an item chosen in the cart. The primary goal is upsell revenue net of refunds, which means waiting for refunds before you read the result.
Measure the take rate (the share of buyers who accept) as a diagnostic, and refunds and cancellations as guardrails. A high take rate with high refunds is not a win.
Mistakes to avoid
- Placing checkout blocks after the freeze. It is a checkout change. Do it before 16 November or not at all.
- A delivery promise you can't keep in every market. A missed promise in December costs more than the extra orders in November.
- Calling a checkout winner on step completion. Shoppers moving faster through shipping is not the same as more orders.
- Reading thank-you codes in December. The window hasn't closed. Read them when it has.
- Counting upsell revenue before refunds arrive.
Checklist
- Your plan checked: Plus for checkout blocks; any plan but Starter for the thank-you page.
- Blocks placed in the checkout editor and checked on a phone.
- One primary goal per test, chosen before launch.
- Buy it now and accelerated checkout handled the same way in every version.
- Checkout tests end by 12 November; blocks kept or removed by 13 November.
- Thank-you codes valid for a set window, with the read date in your calendar.
- Upsell app approval confirmed for your live store.
Running checkout tests for Plus clients? The agency track on the BFCM hub covers checkout tests for Plus clients. A small store on a standard plan? See The thank-you page in Black Friday week. The full season plan is in What to test before Black Friday and Cyber Monday on Shopify.