Ask what powers the "related products" row on a Shopify product page and the documentation gives a more interesting answer than most merchants expect.

Shopify's help center lists three strategies behind auto-generated related products, each with its own requirement. One line stands out:

Product description: Only available to merchants with an English storefront.

Description-based recommendations only run on English storefronts. If your store sells in Japanese, German, or Portuguese, that strategy is off the table — two strategies remain.

And a second structural fact: complementary products — the "frequently bought together" slot — aren't auto-generated at all. You select them manually, capped at ten per product, one product at a time.

Below: exactly where Shopify's native recommendations stop, why they sometimes don't appear at all, and what the app tier costs. Researched September 1, 2026.

The short version

Small catalog, limited sales history? Skip the app. Use the free Shopify Search & Discovery app to set complementary products by hand. Most stores haven't exhausted the native capability before shopping for an app.

Hundreds of products and manual setup isn't viable? There's a native escape hatch: bulk-edit the recommendation metafields via the Bulk Editor or Shopify Flow. Changes sync back to Search & Discovery. Still free.

Want "frequently bought together" derived from actual purchase data? That's app territory. Frequently Bought Together by Code Black Belt is the practical entry point — $14.99/month up to 50 orders, with a free tier for testing.

Want site-wide personalization? LimeSpot AI Bundles & Upsells and similar, where the billing denominator becomes order count or revenue. Note that LimeSpot's Turbo plan is explicitly marked as not supporting Shopify Plus.

All three apps covered here display "This app is not translated into Japanese" on their listings — worth flagging for any non-English-operating team.

Where native stops

Related and complementary are different things

Shopify has two recommendation types with fundamentally different behavior.

Related products are items similar to the one being viewed — potential substitutes, as the help center frames it. These are auto-generated.

Complementary products are items often bought alongside. This is the "frequently bought together" surface. These are not auto-generated.

That asymmetry is the first thing to internalize. Similar products appear on their own. Cross-sells appear only if you decide what they are.

The three strategies, and what non-English stores lose

The help center enumerates the auto-generation strategies with their conditions:

Strategy	Requirement
Purchase history	The product needs previous sales to assess buying behavior
Product description	Only available to merchants with an English storefront
Related collections	Used when purchase history and product description recommendations aren't available

For a non-English storefront, two remain. And since purchase history requires the product to have already sold, a new store — or any product that hasn't started moving — is running on related collections alone.

The practical implication is direct: on a non-English storefront, your collection architecture is your recommendation quality. Writing richer product descriptions won't feed the recommender, because that strategy isn't available to you. Restructuring collections will.

Two further constraints. Auto-generated recommendations can't be edited — you can add your own manually, but you can't correct Shopify's. And they change frequently, which is why the help center warns that what you preview inside the Search & Discovery app will often differ from what's live on your storefront.

Ten per product, one product at a time

The documented workflow: open Search & Discovery, pick a product, select up to 10 complementary products, select up to 10 related products, and choose whether the manual list displays alone or alongside auto-generated results.

Ten is generous enough for most catalogs. "One product at a time" is the actual problem. Five hundred products means five hundred passes.

The native answer is metafield bulk editing. Select multiple products on the Recommendations page, open the Bulk Editor, and set recommendations in bulk. The help center also notes you can edit these metafields from other apps, including Shopify Flow.

Three standard metafields drive it:

• Related products — your manually selected list. Does not include Shopify-generated recommendations
• Related products settings — accepts only manual (show only your selections) or ahead (show your selections before auto-generated ones)
• Complementary products — your manually selected cross-sells

Edits sync with the Search & Discovery app. Bulk recommendation management is therefore achievable without buying anything.

Six reasons a recommendation doesn't show

This is where "I set it and nothing happened" tickets get resolved. A recommended product displays only if:

• It isn't sold out (related products can still show if continue selling when out of stock is on; complementary products require inventory above 0)
• It doesn't have Unlisted status
• Its price is above $0.00
• It isn't a gift card
• It's published to the Online Store sales channel
• It isn't currently in the visitor's cart

That last one causes real confusion during testing. Add a product to your own cart, reload the product page, and it vanishes from the recommendations. Working as designed.

The inventory asymmetry catches people too. A product with "continue selling when out of stock" enabled will appear as a related product but never as a complementary one. If your plan was to cross-sell a preorder item, that plan doesn't work.

Theme requirements

Configuration alone doesn't render anything. The help center is explicit: to display recommendations, you need a theme that includes a complementary products section and a product recommendations section, and you have to add those sections to your product pages.

Complementary products is a block available only in compatible themes, and only one complementary products block is allowed per product page.

The block's own settings are bounded: maximum products to show is 1–10, products per slider page is 1–4, pagination style is dots / counter / numbers, and image ratio is portrait or square.

On the developer side, the Product Recommendations API mirrors this: limit accepts 1 to 10 with a default of 10, and intent accepts related or complementary, defaulting to related. The ten-item ceiling is an API property, not a theme choice.

Combined listings

Stores using Combined Listings get an extra setting: in Search & Discovery's settings, choose whether auto-generated recommendations show only child products (the default), only parent products, or both. The setting only appears if your store contains combined listing products or has the Combined Listings app installed.

When an app is actually warranted

Three cases:

You want cross-sells derived from real purchase data. Complementary products are manual-only natively. Reflecting "what actually got bought together" requires something computing it.

You want recommendations outside the product page. Native recommendations are product-page sections and blocks. Cart drawers, thank-you pages, and checkout are elsewhere.

You want per-visitor personalization. Native auto-generation operates per product, not per visitor. There's no browsing-history-driven variation.

If you only want "similar products" and "buy these together" on the product page, native plus metafield bulk editing covers it.

Three recommendation apps

Frequently Bought Together (Code Black Belt)

Built for Shopify badge, 4.8 across 1,045 reviews (989 five-star, 11 one-star). The listing notes the app was created in 2017. Supported languages: English, French, German, Spanish, Italian, Dutch, Brazilian Portuguese, Simplified and Traditional Chinese — not Japanese.

It renders a "frequently bought together" bundle on the product page with one-click add-to-cart for the set. AI-driven automatic recommendations are supplemented by per-product manual tuning, and four bundle discount types are available.

Billing denominator: monthly order count.

Plan	Monthly	Terms
Starter	Free	Up to 3 bundles, basic widget settings, basic support
Professional I	$14.99	Up to 50 orders/month. Unlimited bundles, auto recommendations, bundle discounts, analytics, translation
Professional II	$19.99	Up to 500 orders/month
Professional III	$39.99	Over 500 orders/month

Professional tiers include a 14-day trial. The free Starter tier is positioned for testing basics — three bundles, and no automatic recommendations, which are a Professional feature.

LimeSpot AI Bundles & Upsells (LimeSpot)

4.6 across 391 reviews. English only; not translated into Japanese. It places recommendations and offers across the homepage, product pages, cart, checkout, and post-purchase, and lists integrations including Shopify Flow, Klaviyo, LoyaltyLion, and Okendo.

Two different billing denominators, depending on plan.

Turbo bills on order count. Free up to 10 orders/month; beyond that it starts at $9.99/month and tiers upward, with the listing citing "$19.99/month up to 250 orders" as an example. 15-day trial.

Turbo also carries a notable restriction, attached to both the free and paid tiers: it does not support Shopify Plus.

Max bills on store revenue, starting at $50/month, with the listing citing "$150/month up to $50k revenue." It adds audience segmentation and journey funnels, personalized bundle discounts, smart collections, email and SMS personalization, A/B/n testing, and 70+ app integrations.

Because the denominator switches from orders to revenue between plans, model your cost curve before you grow into it.

Product Recommendations & Upsell • ES (EcomStar)

Built for Shopify badge, 5.0 across 130 reviews (all five-star). English only; not translated into Japanese. Places frequently-bought-together and related product offers on product pages, cart, and thank-you page, with GemPages and PageFly integrations listed.

Two tiers on monthly order count. The "Free Forever" plan includes all upsell and cross-sell features, theme-matched styling, and Zoom/chat/email support. Basic is $9.99/month with the first 50 orders per month free and overage billed beyond that. The trial is 20 days — the longest of the three.

One review from May 26, 2026 describes support explaining how to use the app in conjunction with Shopify's Search & Discovery app, which is a useful reminder that native and app aren't mutually exclusive.

Side by side

	Frequently Bought Together	LimeSpot AI Bundles & Upsells	Product Recommendations & Upsell • ES
Developer	Code Black Belt	LimeSpot	EcomStar
Rating / reviews	4.8 / 1,045	4.6 / 391	5.0 / 130
Built for Shopify	Yes	Not listed	Yes
Free plan	Yes (3 bundles max)	Yes (10 orders/mo)	Yes (Free Forever)
Paid entry point	$14.99/mo	$9.99/mo	$9.99/mo
Billing denominator	Monthly orders	Orders or revenue	Monthly orders
Trial	14 days	15 days	20 days
Japanese UI	No	No	No
Plus support	Not stated	Turbo explicitly unsupported	Not stated

Note that the three "free plans" are constrained along completely different axes: bundle count, order count, and — in ES's case — a Free Forever tier alongside a paid plan whose first 50 orders are also free. "Has a free plan" is not a comparable data point here.

Choosing by situation

Under 100 products, and you want deliberate cross-sells

No app needed. Set complementary products manually, ten per product. At this scale, human judgment usually beats the algorithm. The help center even offers guidance: pick add-ons priced lower than or equal to the original product.

Large catalog, manual setup impractical

Try metafield bulk editing first — Bulk Editor or Shopify Flow against Complementary products and Related products. If that works for you, there's no monthly cost.

You want cross-sells from purchase data

Frequently Bought Together is the first candidate. At 50 orders/month or fewer, that's $14.99/month, backed by over a thousand reviews. English admin only.

You want recommendations beyond the product page

LimeSpot or ES. If you're on Shopify Plus, check LimeSpot's Turbo exclusion before you start a trial.

Recommendations aren't appearing

Work the six display conditions before evaluating apps. "Not in the visitor's cart" and "complementary requires inventory above 0" are the two most often mistaken for bugs. Then confirm your theme actually has the complementary products block.

Before you install

• Have you set complementary products at all, knowing they're never auto-generated?
• If your storefront isn't English, do you recognize that collection structure drives recommendation quality?
• Have you counted how many products need manual setup and evaluated metafield bulk editing?
• Does your theme include the complementary products block and product recommendations section?
• Have you walked the six display conditions?
• Do you know whether the app you're considering bills on orders, revenue, or products?
• On Shopify Plus, have you checked for excluded plans?
• Have you evaluated UI language separately from support language?

The bottom line

Four findings, verified September 1, 2026.

Only related products are auto-generated. Complementary products stay empty until you fill them, ten per product.

Of the three auto-generation strategies, product-description-based recommendation is documented as English-storefront-only. Non-English stores run on purchase history and related collections — and for products without sales, collections alone.

Six display conditions gate whether a recommendation renders. Two of them — "not in the visitor's cart" and "complementary requires inventory above 0" — routinely get mistaken for defects.

The major recommendation apps bill on order count or revenue, and their free tiers are constrained along non-comparable axes.

Start by opening Search & Discovery and checking whether you've set a single complementary product. If the answer is zero, there's free capability to use before anything gets a monthly invoice.

Pricing and specifications change. This article reflects the Shopify Help Center, Shopify.dev, and Shopify App Store as of September 1, 2026. Verify current details before adopting.