Checkout upsells for BFCM 2026: what our Shopify Plus data says actually sells

Every October I end up having some version of the same conversation with merchants. What should we put in checkout for Black Friday? And for a while my answer was a bit of gut feel, backed by whichever store I happened to be looking at that week.

This year I wanted a better answer, so I went digging through our data. I looked at every completed checkout that showed an upsell between 20 July and 4 October 2026, and at last year's Black Friday for the stores we can follow all the way from October into Black Friday week.

The short version, if you've only got a minute. During a sale, shoppers take the cheap thing. Last Black Friday week the median upsell was 11.8% of the order, down from 18.9% in October. Offers priced under 10% of the cart got 0.96 adds per 100 times they were shown, while offers priced above the entire cart got 0.34. And first-time customers took upsells more often than returning ones, 2.25% against 1.91%, which matters because 52.1% of the checkouts that showed an upsell were someone's first order, and that's in a normal week.

If you run ecommerce at a brand, I hope this is useful while you set checkout up before the traffic arrives. If you're on the agency or app side, I'm keen to hear whether it matches what you're seeing.

A pinch of context

Everything here comes from Shopify Plus stores running checkout upsells through Checkout Components, which is the app I've been building since 2023. So yes, it's our data, and I've tried to call out where that skews things.

The main window is 20 July to 4 October 2026.

A few terms I'll use a lot. Attach rate is the share of completed orders that include something added through a checkout upsell. Take-up is how often an offer gets added when it's shown. When I say an upsell was "15% of the order", I mean the upsell item's price divided by the order subtotal, with the upsell included in that subtotal.

Wherever a table shows a median, it's the median store, so one huge brand can't drag the number around for everyone. Where I've pooled every order or checkout together I say so, because pooled numbers lean towards the bigger stores. Post-purchase offers are left out of all of it.

Most of what follows is in percentages, so it reads the same whatever currency a store sells in. Where I do talk dollars, they're US dollars from US-dollar orders only.

So what actually happened last Black Friday?

I took every store that had a steady run of checkout upsells through October 2025 and followed them into Black Friday week, 24 to 30 November. Same stores on both sides, so the change isn't just us adding new merchants.

In Black Friday week, those stores sold 349.1% more upsell items than in an average October week. But the upsells got smaller. The median upsell went from 18.9% of the order in October to 11.8% in Black Friday week.

That drop is the bit I keep coming back to. Here's what was getting added:

I've grouped products by their titles. Anything that didn't fit a group cleanly is left out, along with a few categories too small to read anything into, so the shares don't add up to 100%.

Everything sold more, because that's what Black Friday does. What changed was the mix. Socks went from 7.2% of upsells to 20.5%, and beauty from 12.5% to 23.9%. Bags and accessories still sold more than in a normal week, but their share of the mix fell from 15.0% to 5.6%.

My read on it is pretty simple. On Black Friday the shopper has already made the big decision, usually on a discounted hero product. By the time they hit checkout they're done deciding. A cheap pair of socks is an easy yes. A bag is a whole new purchase, and most people aren't up for that with a payment screen in front of them.

One wrinkle worth flagging. On the smaller set of US-dollar orders, the median upsell price was US$19.20 in both periods. So take the drop in size as a pattern across the whole fleet, not something every market showed.

And one honest limitation. We don't have last year's total order counts in the same table, so I can show you volume and mix but not last year's attach rate. That's what June is for.

June was the dress rehearsal

This year we do have total orders, and the second half of June had a big sale period, which makes it the closest thing to a practice run for BFCM. I followed a fixed group of stores that had upsells live for at least 30 of the 35 weeks from February to early October.

The pooled attach rate went up during the sale, but the median store's didn't. It sat around 1% the whole way through. So the bigger stores turned sale traffic into more upsells, while the typical store just got more traffic at the same rate. I'd love to tell you exactly why, and I don't think one table can, but my guess is that the bigger brands planned their checkout for the sale and the rest left it as it was.

Upsell size did exactly what it did at Black Friday. It fell from 20.0% of the order in late May to 12.1% at the peak, then came back to 19.5% once the sale was over.

What sells next to what

This was the one I was most excited to pull. For every order with an accepted upsell, I looked at the most expensive thing already in the cart and compared it with what got added at checkout.

The "different category" column leaves out shipping protection and gift wrap, which are their own thing and much smaller. Food and supplements would have made this table too, but one store makes up 41.5% of that sample, so I've left it out rather than publish one merchant dressed up as a trend.

A few things worth knowing before you set up your mappings:

• →When someone buying clothing takes a second piece of clothing, it's a real add, 26.2% of their order. Cross-sells off a clothing cart come in at 12.6%. If you're a fashion brand, the "complete the look" upsell is earning its keep.
• →Shoe carts barely take more shoes. 56.4% of the upsells on footwear carts were socks. If you sell shoes and you're not putting socks in checkout, I'm not sure what you're doing.
• →Beauty breaks the pattern. Beauty shoppers mostly stick to beauty, but when they do cross over, the add is bigger than the same-category one.
• →Home & kitchen carts stay in home & kitchen 59.2% of the time.

Offer price vs what's already in the cart

Our upsell events record every product shown and what the cart was worth at that moment, so I could line up the price of each offer against the cart it was shown in.

The line goes down at every step, which isn't a shock. What did surprise me is how many offers sit in the bottom two rows. 37.2% of all offers shown cost more than half of what the shopper was already spending, and 17.0% cost more than the whole cart.

I think this happens because most upsells are set up once, against an average cart, and then shown to everyone. An offer at US$60 is fine next to a US$300 order. Next to a US$45 order it's asking someone to more than double their spend in the last step.

These are adds, not dollars. A pricier offer that gets taken less often can still bring in more revenue per view, so look at revenue per impression on your own store before you cut every expensive offer.

For what it's worth, the upsells people actually bought had a median of 15.8% of the order, and 67.9% of them cost less than a quarter of it. On US-dollar orders, the median accepted upsell was US$22.40.

Who actually takes an upsell

This is every completed checkout in the window that showed an upsell, pooled. The attach rate here is the share of those checkouts where the shopper bought something from the upsell.

I went in expecting loyal customers to be the ones adding extras. It's the other way round. First-time buyers took upsells more often than returning ones, and guests more often than people who'd logged in.

That has a very practical consequence for BFCM. If you've set your best offer to show only to VIPs or logged-in customers, you've hidden it from the people most likely to say yes. In our window, 85.2% of those checkouts were guests and 78.9% were on a phone, and that's in a normal week.

The discount code result is a nice one to have, too. Shoppers using a code took upsells at the same rate as everyone else. So the idea that bargain hunters won't add anything doesn't hold up, at least not in a normal week.

Cart size was the clearest split. Shoppers with 3 or more items took an upsell at 2.65%, against 1.65% for single-item carts. People who are already browsing and building a basket keep going.

What each vertical is running

I put each store into a vertical based on what its shoppers were buying, then looked at which upsell formats got shown and how often an impression turned into a sale. The format columns are the share of that vertical's upsell impressions.

The last column is pooled, so the bigger stores in each vertical count for more. Baby and kids had too few stores to include.

Fashion is where most of our merchants are, and it converts lowest per impression. Beauty converts best, which lines up with what we saw above. Beauty shoppers love adding another beauty product.

Format-wise, the plain upsell list does most of the work everywhere. Home & kitchen leans on carousels far more than anyone else, which makes sense when the product needs a bigger picture.

The format I'd pay attention to is the checkbox. At the median store, 2.13% of checkbox impressions ended in a sale, against 0.86% to 0.95% for lists, carousels and free-shipping bars. Fewer stores run checkboxes than any other format, so treat that with some care. But checkboxes hold cheap, one-tap add-ons, and cheap add-ons are what BFCM shoppers were reaching for last year. Food & supplements is the only vertical giving it real room.

One more thing on where the products come from. Hand-picked product mappings had the best median, 1.14% per impression, and also the widest spread. The best quarter of stores using them hit 3.01% or more, and the worst quarter sat at 0.39% or less. Shopify's automatic recommendations were steadier, at a median of 0.94%. A mapping is only as good as the last time someone looked at it.

Where your store probably sits right now

If you want a benchmark to hold yourself against before the sale, here it is. These are stores that processed at least 500 orders in the window and had an upsell live for at least 60 of its 77 days. Attach rate here is measured against every completed order, including the ones that never saw an offer.

So a typical store gets an upsell into just under 1 in 100 orders. If you're under 0.45%, you're in the bottom quarter.

I split those stores into quarters to see what the top ones had in common.

The biggest gap is coverage. The typical bottom-quarter store logs fewer upsell impressions than it has orders, so some of its checkouts never see an offer at all. That's not proof the top stores win because of it. Bigger brands have more staff and more products to play with. But every row in that table is something you can change in your own checkout settings this week.

My BFCM checklist

If I were setting up a Plus checkout for Black Friday this week, this is what I'd do, in this order.

1. Lead with something cheap. Make your first offer cost under 10% of a typical cart. Socks, minis, samples, shipping protection. That's where the mix moved last Black Friday.
2. Put a cart value rule on anything expensive. If an offer costs US$60, there's no reason to show it on a US$45 cart. Offers priced above the whole cart got 0.34 adds per 100 views.
3. Match the offer to the cart. Clothing with clothing, shoes with socks. If you sell beauty, try one offer from outside your core range and see what happens.
4. Show your best offer to guests. Don't hide it behind a VIP tag or a login. First orders and guests took upsells more often than anyone else.
5. Check your coverage now, before the traffic comes. If your store logs fewer upsell impressions than orders, some checkouts aren't seeing anything. Start with your display rules.
6. Add a checkbox. One low-cost add-on, ticked by the shopper, sitting under everything else.
7. Break your free gift on purpose. Add the gift, remove the item that qualified for it, and see if the gift is still free. Better you find it than a few thousand bargain hunters.
8. Lock it in early. Run any tests in October and set your baseline. Black Friday week is a terrible time to find out which variant wins.

If you want a second pair of eyes on it, our free checkout audit comes back within 5 minutes.

What this data can't tell you

I'd rather you trust these numbers a little less than more, so here's where they're weak.

They only come from stores running Checkout Components. A merchant paying for an upsell app is probably more into upselling than the average Plus store, so if anything these numbers flatter the category.

Stores in the data sell in several markets and currencies. That's why most figures are percentages, and why the dollar figures come only from US-dollar orders, which are a smaller slice of the whole.

Some orders that go through express checkout wallets don't get tied back to the checkout session that showed the upsell. That undercounts attach, so treat every rate here as a floor.

And the vertical, product and format splits are correlations. Stores that do well tend to share habits. That doesn't mean copying the habits will move your number, which is why I keep banging on about testing.

We'll run all of this again once BFCM 2026 is done and I'll publish what changed. If you've spotted something I've read wrong, or you've seen something different on your own stores, let me know. I'd genuinely like to hear it.