Black Friday changes that quietly cost small stores sales

• Don't publish a new theme in the last two weeks. Duplicate the live one and make small, checked edits instead.
• Every app you add for the sale is new code on every page. Add them before the freeze, or skip them this year.
• Your banner, product page and checkout must promise the same offer. If they disagree, shoppers find out at the payment step.
• Check every sale page on a phone, and place a test order with each discount code you'll share.
• Decide how your discounts combine, deliberately. Codes stack only when you allow it, and stacked discounts can cost more than you planned.

The freeze itself is covered in part 3 of this series, and the full calendar in our general Black Friday testing guide.

A new theme is the biggest change you can make to a store. It changes every page, the cart, how apps attach to the page and how your tracking loads. Publishing one the week before Black Friday leaves no time to find what broke.

Instead: duplicate your live theme in the Shopify admin and keep the copy as a backup. Make sale edits (a banner, a collection) on the live theme in small steps, and check each one on a phone. Save the redesign for January.

Countdown timers, spin-to-win popups, extra upsell widgets and chat buttons all add scripts. Shopify's own guidance on web performance notes that each feature you add to your online store can affect its performance, and some can slow the site down or cause unexpected behavior.

What to check for each new app:

• Does it slow the first screen on a phone? Compare before and after on mobile data.
• Does it cover the Add to cart button on a small screen?
• Does it show a price or offer of its own? A popup promising 15% off while the banner promises 20% creates a question you can't answer at checkout.
• Can you turn it off quickly if it misbehaves during the sale?

If an app isn't in place and checked by 15 November, leave it out.

The shopper reads "20% off everything" on the banner, adds three items, and checkout applies 20% to two of them because one is excluded. Or the banner says "free shipping" and checkout charges shipping below a threshold the banner didn't mention. That surprise lands at the payment step, where Baymard's research finds "extra costs too high" is the most common reason for abandoning (40% of shoppers who were not just browsing, page updated 22 September 2025).

Instead: write the offer once, in one place, with its conditions ("20% off, excludes gift cards", "free shipping over $60 after discounts"), and copy the same words to the banner, the product page, your emails and the discount's title in Shopify. Then place a test order that matches the banner exactly. Know which cart total the threshold counts: a Shopify free-shipping discount counts prices after discounts that combine with it, while a Benchmyrk shipping test counts the cart before discounts.

Most small stores check changes on the computer they built them on. Sale traffic from social posts and emails is often mostly phones. A banner that wraps to four lines on a phone can push the product photo and the button below the first screen. Baymard also lists website errors or crashes as a reason for 17% of abandoned purchases (same page).

Instead: for every page linked from a sale email or post, open it on your own phone, on mobile data, and go as far as the payment step.

Shopify discounts belong to three classes: product, order and shipping. They combine only when each discount is set to combine with the other's class, and product discounts apply first, before the order subtotal is calculated. That's flexible, and it's easy to get wrong.

How to avoid it:

1. 1.List every discount that will be active from 26 to 30 November: the sale, welcome codes, loyalty codes, influencer codes, free shipping. Shopify allows up to 25 active automatic discounts per store, app discounts included (Shopify Help Center, read 11 October 2026). Benchmyrk's price, offer and shipping tests use up to four of them. A late app discount can push you over the limit.
2. 2.For each pair, decide whether it should combine, and set it in each discount's combinations settings.
3. 3.Place a test order for each combination a real shopper might try.
4. 4.Give every sale discount an end date.

Something will go wrong during the sale: an out-of-stock product, a typo, a slow page. Fix the problem directly (hide the product, correct the text, remove the app). Don't respond by redesigning the page or adding another app; that's a second untested change on your busiest day.

You can't prove a small loss at small-store traffic, but you can look for breakages:

• The funnel by device. If checkout completion on phones drops sharply on a day you changed something, look at that change.
• Messages from shoppers. "The code didn't work" or "the banner said free shipping" are free bug reports.
• Your test orders. If one fails, many shoppers' orders did too.

• A new theme in November. Duplicate and edit instead.
• New apps after 15 November. Each one is new code on every page.
• Two different offers on one page. Banner, popup and checkout must agree.
• Checking on a computer only. Your sale traffic is on phones.
• Codes that combine by accident. Set combinations deliberately and test them.

• [ ] Live theme duplicated, no new theme published in November
• [ ] Every sale app installed and checked on a phone by 15 November
• [ ] Offer written once, with conditions, and copied everywhere
• [ ] Every active discount listed, with its combinations set
• [ ] Test order placed for each realistic combination
• [ ] Every page in sale emails checked on a phone on 25 November
• [ ] A plan for fixing problems during the sale without redesigning

Managing several client stores? The agency track on the BFCM hub covers what not to test over BFCM. There's also an overview of app impact testing.