Section Store Help CenterOpen Section Store

How do recommended products in the cart drawer work, and why are they expensive?

AI recommendations come from Shopify. We add no price filter. See what each side controls and when to use manual rules.

Updated November 4, 2026

AI recommendations come from Shopify. The drawer sends Shopify one product, the last one in the cart list, and shows what comes back. We don't filter the picks by price or cart value, so a pick can cost more than the whole cart. There's no price limit. To choose exactly what shows, switch Recommend by to Manual rules, add a rule, pick your products and click Save.

Which product the drawer sends

The drawer sends the last product in the cart's list. That is the lowest product line in the drawer. It skips free gifts, order addons and products sold by another store.

Shopify lists the newest product first, so this is usually the first product the shopper added. You can't choose a different product. The picks relate to that one product, not to the whole cart.

What we control

  • Which product we send. See above.
  • Which picks we ask for. That is your Auto recommendation type. If Complementary products returns nothing, we ask for Related products.
  • What we drop. We drop sold-out products. With Hide products already in cart on, we drop those too.
  • How many show. That is your Recommendation limit, from 1 to 10.

We don't use price or the cart total at any step. So a cheap product can bring a costly pick.

What Shopify controls

  • Which products come back, and their order. We show them in Shopify's order.
  • Related products. Shopify builds these from sales data, product descriptions and collections.
  • Complementary products. Shopify doesn't build these. You set them up in the Shopify Search & Discovery app.
  • How many come back. Shopify returns up to 10.

What you can change on AI recommendations

  1. In Shopify admin, open Apps → Section Store → Cart Drawer.
  2. Click Edit on your cart.
  3. Under Body, open the Products upsell card.
  4. Set Auto recommendation type to Complementary products or Related products.
  5. Lower Recommendation limit to show fewer picks.
  6. Tick Hide products already in cart if you want that.
  7. Click Save.

To change the picks themselves, open the Shopify Search & Discovery app. It manages related and complementary products for your store.

Take full control with Manual rules

  1. In Shopify admin, open Apps → Section Store → Cart Drawer.
  2. Click Edit on your cart.
  3. Under Body, open the Products upsell card.
  4. Set Recommend by to Manual rules.
  5. Click Add rule.
  6. Set When cart contains to Any product.
  7. Under Recommend these products, click Add product and pick your lower-priced add-ons. A rule holds up to 5.
  8. Click Save.

Shoppers now see only the products you picked. You can add rules for selected products or collections above that one. The first rule that matches wins.

When to switch to Manual rules

  • You want picks in a price range.
  • You sell a few add-ons that suit most orders.
  • You want set pairings, like a case with a phone.
  • Your store is new and Shopify shows few picks.

Keep AI recommendations if you have a large catalog and don't want to maintain rules.

What you can't set in the app

  • A price limit or a price sort on AI recommendations.
  • A list of products to keep out of AI recommendations.
  • A rule based on cart value.
  • A discount on upsells.

Also searched as: upsell too expensive, expensive recommendations, upsell costs more than cart, recommended products too pricey, wrong products recommended, irrelevant upsell, how recommendations work, related products, search and discovery app, price limit on upsell, ai recommendations price, which product recommendations use, upsell based on whole cart, exclude product from recommendations, expensive upsale, cart drawer recommended products, how do recommended products work, control recommended products, recommended products in cart drawer, recommends high value items, recommends expensive items, max price on upsell

Did this answer your question?