The section below the product page that lists "Recommended for You." Can you tell if it's an app or a native feature?

The answer is a native feature. Shopify has a Product Recommendations API that is built-in, and adding a related products section to a theme will make it work without any additional cost. Even without the Shopify Search & Discovery app, automatically generated related products will be displayed.

The issue is where the limits of this native feature lie. The official documentation clearly states these limits. The return limit is 10 items. Intent is limited to two values. Cross-sells are not automatically generated. And Japanese stores have a structural weakness that English stores do not.

This article outlines the four boundaries where a recommendation app becomes worth the cost, based on the official documentation and verified on September 1, 2026. All prices are in USD.

Conclusion

If your goal is product page recommendations with sufficient order data and well-designed collections, an app is not necessary. The native feature is sufficient, and there is no additional cost.

If you want to offer cross-sells or control the content of the recommendations, you can do so manually with the free Shopify Search & Discovery. You can set up to 10 items per product.

If you have a large number of products and manual setup is impractical, or you want to add bundle discounts, consider a paid app. For stores with up to 50 monthly orders and a need for Japanese UI, consider Wiser AI - Upsell & Cross Sell (free tier for up to 50 orders), for stores with up to 500 monthly orders and a focus on bundle discounts, consider Frequently Bought Together CBB ($14.99+), and for stores with over 500 monthly orders, consider Also Bought CBB ($19.99+).

If you want to display, personalize, A/B test, or track sales from recommendations on the cart, checkout, thank you page, or email – these are the four scenarios where you should consider moving away from the native feature.

Verify the Native Feature's Specifications with Numbers

The API Returns at Most 10 Items

The Ajax API's /{locale}/recommendations/products.json has a limit parameter. The official reference states that the range is 1 to 10, with a default of 10.

The Storefront API's productRecommendations does not have a limit argument, but it is defined to return "up to 10 Product objects." Therefore, neither API has an 11th item for each product.

Shopify's implementation guidelines are even more conservative. For related products, they recommend "limiting to the top 10 most relevant items per product, and capping at 4 items per product page." For cross-sells, they recommend "defaulting to 2-3 items, with additional items paginated."

Intent is Limited to Two Values

The intent parameter can only accept the values related (related products) and complementary (cross-sells), with a default of related. Passing any other value will return a 422 error, and the error message will state "The intent parameter must be one of related, complementary."

This two-value structure was added in the changelog on September 20, 2022, and has not changed since.

Cross-Sells Are Not Automatically Generated

This is a common misunderstanding. The official documentation clearly states that "Shopify automatically generates related products, but not cross-sells. Cross-sells require manual setup."

To offer cross-sells, you can set them up manually in the Shopify Search & Discovery app. You can set up to 10 items per product. You can also manually set related products, with a maximum of 10 items, and choose to display them either manually or automatically.

Additionally, there are conditions on the display side. Cross-sells must be added as a block and are only available in compatible themes. You can add one block per product page.

Seven Conditions to Exclude from Recommendations

The Help Center lists the following conditions for recommended products:

• Not out of stock
• Not unpublished
• Has a price above zero
• Not a gift card
• Is published on an online store channel
• Not currently in the visitor's cart
• Is marked as active.

The only difference in handling stock is between related and cross-sells. Related products are shown even if the stock is set to "Continue Selling," but cross-sells require a stock count above zero. Many issues with cross-sells not showing are due to this.

Japanese Stores Lack One Algorithm

This is the most important part of the article.

There are three strategies for automatic product recommendations. Purchase history (items purchased together in the past), product descriptions (items with similar descriptions), and related collections (items within the same collection, as a fallback).

The Help Center outlines the prerequisites for each strategy. For product descriptions, it states:

Product descriptions: Only merchants with an English storefront can use this strategy.

This is the same on the Japanese version of the help page. Therefore, Japanese stores cannot use one of the three strategies from the start.

This limitation applies when:

• New products or stores with no sales history cannot use purchase history. Since Japanese stores cannot use product descriptions either, related collections are the only basis for recommendations. This often results in a situation where only other items from the same collection are shown.
• Additionally, the official documentation mentions another overlooked limitation. Orders imported from other stores or other e-commerce platforms are not used for recommendation generation. Even if a store has purchase data, the purchase history strategy starts from scratch.

For newly imported Japanese stores, both of these apply. Expecting high accuracy from standard recommendations from the start is unrealistic. Operating with collection design and manual settings for a while is more practical.

Note that the official documentation also mentions some caveats about the generated results. "The recommendations for a product are frequently changed," and "The previews in the Search & Discovery app often differ from the display on the storefront." The preview and live display can be different by design.

Effect Measurement is Possible, but Sales Figures Are Not Included

This is not widely known, but the effectiveness of product recommendations can be measured with standard reports. There are two of them.

Product recommendation conversions over time shows daily session numbers, click sessions, add-to-cart sessions, purchase sessions, and click, add-to-cart, and purchase rates.

Product recommendations with low engagement extracts items with a click rate below the average from the top-selling products in the last 30 days.

However, these two reports do not include sales figures. Only session numbers and rates are shown. To track how much was sold through recommendations by amount, the standard feature does not provide an answer.

There are some prerequisites for measurement. The theme must use the API-specified URL parameters (such as pr_choice, pr_prod_strat, etc.), and cart additions and purchases through Buy Button or third-party apps are not included in the reports. If you are using the app to generate recommendations, nothing will appear in the standard reports.

Shopify Search & Discovery is Free, but Rated at 2.7

The free app for overwriting or adding to the standard recommendations is Shopify Search & Discovery, developed by Shopify itself, with a free tier and support for the Japanese language.

You can manually add related products (up to 10 items), set cross-sells (up to 10 items), control the display order of manual and automatic recommendations, and hide automatic recommendations for individual products. You can also differentiate parent and child combinations. Settings are done through metafields, so you can also use bulk editing tools or Shopify Flow.

You need to enable "Shopify Network Intelligence" in your privacy settings.

However, the app's rating on the App Store as of September 1, 2026, is 2.7 (464 reviews), with 19% (90 reviews) giving it one star. While recent reviews indicate dissatisfaction mainly with the relevance of search results and the filter limit (25 items store-wide), there were no mentions of the recommendation feature itself. It's better to not judge the feature solely by the rating.

Four Boundaries Where an App Becomes Worth the Cost

The limitations of the native feature are not about the fine-tuning of the feature but are categorized into the following four areas.

1. Placement. The Product Recommendations API is a scoped API that requires a product_id. It does not have a mechanism to take in cart contents or session history as input, so cart, checkout, thank you page, and email recommendations are out of the scope of the native feature.

2. Personalization. The only officially listed algorithm inputs are "store-wide purchase history," "product descriptions," and "related collections." The only officially documented behavior that depends on the visitor is "excluding items in the cart," and there is no mention of dividing based on browsing history.

3. A/B testing. This feature is not documented in the Search & Discovery feature list or the Help Center.

4. Measuring sales revenue. As previously mentioned, the standard reports do not include a revenue column.

If none of these apply, there is little reason to pay a monthly fee.

Three Billing Bases — None Charge by Impressions

When comparing recommendation apps, listing only the monthly fees is meaningless because the billing bases differ. Across the six apps we checked, the models fall into three categories.

Application	Developer	Rating/Reviews	Charging Denominator	Price	Japanese UI
Shopify Search & Discovery	Shopify	2.7/464	None (Free)	$0	Yes
Wiser AI - Upsell & Cross Sell	Expert Village Media Technologies	4.9/513	Monthly Orders	0-50 orders = Free / 51-100=$9 / 101-300=$19 / 301-500=$49	Yes
Frequently Bought Together CBB	Code Black Belt	4.8/1,044	Monthly Orders	~50 orders=$14.99 / ~500=$19.99 / Unlimited=$39.99	No
Also Bought CBB	Code Black Belt	4.8/314	Monthly Fixed	$19.99 (Unlimited items, orders, and traffic)	No
LimeSpot AI Bundles & Upsells	LimeSpot	4.6/391	Turbo is monthly orders / Max is store sales	Turbo $9.99~ (Up to 250 orders=$19.99) / Max $50~ (Up to $50k sales=$150)	No
Rebuy Personalization Engine	Rebuy	4.7/748	Monthly Orders (OPM)	Package $25/month ~, Platform One $534/month ~	No

None of the apps in this review charge by impressions. Assuming that recommendation apps use impression-based pricing leads to the wrong comparison.

Three points can be read from this table.

There are calculable break-even points for order-based and fixed-fee charging. Also Bought CBB is a full fixed fee of $19.99. The same developer's Frequently Bought Together CBB charges $19.99 up to 500 orders, and above that, it charges $39.99. Even within the same developer's two applications, the fixed-fee model becomes cheaper when exceeding 500 orders. Wiser charges $49 for 301-500 orders, so the break-even point is between 300 orders (Wiser $19) and 301-500 orders (Also Bought $19.99).

Order-based charging does not consider the value of a single order. Therefore, stores with fewer orders and higher average order values (e.g., 80 orders/month, $30,000 per customer) are more advantageous with order-based charging, while stores with more orders and lower average order values (e.g., 2,000 orders/month, $2,000 per customer) are more advantageous with fixed fees.

Revenue-based charging increases fixed costs as success increases. LimeSpot Max charges $150/month for 30 days of sales of $50,000. Note that the developer has publicly replied to a review about the billing amount, stating that it is "2.5% of driven revenue," which demonstrates that revenue-based charging makes it difficult to predict the billing amount. This is an individual review and the developer's response, not a general evaluation, but it provides insight into the nature of the denominator.

Two Constraints on Plan Limits

LimeSpot Turbo is not available for Shopify Plus stores. Both the App Store and the developer's official website clearly state "Turbo is not available for Shopify Plus stores." Plus stores automatically use the revenue-based Max plan ($50/month ~). If you plan to sign up for a cheaper plan on Plus, the premise will be broken here.

Rebuy's Checkout Extensions are exclusive to Shopify Plus. The official pricing page states "Shopify Plus exclusive." If the purpose is to display recommendations within the checkout, the first branch is whether the store is Plus or not.

Two Options if Japanese UI is Needed

Only two applications have Japanese UI: Shopify Search & Discovery and Wiser AI - Upsell & Cross Sell. LimeSpot and Rebuy are in English only, and the two Code Black Belt applications do not have Japanese as a language option. Nosto is in English, French, and German.

Note that Japanese UI and Japanese support are separate issues. In this investigation, we could not confirm the presence or absence of Japanese support for any of the applications.

For Wiser AI, there is an additional note. The free plan (0-50 orders/month) and paid plans have identical features, with the only difference being the order range. For stores with 50 or fewer orders per month, you can try the product page, cart drawer, thank you page, and post-purchase upsell for free. We could not confirm the pricing for plans exceeding 500 orders, as it is not listed on the App Store page and requires referencing an external page.

For Nosto, the pricing is completely undisclosed. The App Store states "Free to install. Additional charges may apply." It also explicitly states "An Nosto account is required, and it is not free." Since we cannot include numerical values in the table, consider an estimate when evaluating.

Common Misunderstandings

Misunderstanding	Reality
You need an app to display related products	It is a native feature. You can add sections to the theme.
You need Search & Discovery to display related products	Automatic generation works without the app. The app is for overwriting or adding.
Complementary items are displayed automatically	Complementary items are not automatically generated. Manual settings are required.
Increasing the limit allows you to display 20 items	The limit is 10 items. The Storefront API does not have a limit argument.
The algorithm is the same for Japanese stores	Product description strategies are limited to English storefronts. Japanese stores have one less.
Past orders carry over when moving from another cart	Imported orders are not used for recommendation generation.
You cannot measure the effectiveness of recommendations by default	There are two standard reports. However, there is no column for revenue.
The Search & Discovery preview shows on the storefront	It is explicitly stated that it is often different.
Recommendation apps are charged by impressions	The six apps reviewed use order-based, flat-rate, or revenue-based pricing.
LimeSpot is affordable on Plus	Turbo is not available for Shopify Plus stores.

Pre-Installation Checklist

• Confirm whether related products are displayed by the native feature or an app.
• Distinguish between "not set" and "out of stock" for complementary items.
• Revise your collection design considering that the standard algorithm is one less for Japanese stores, especially for new products.
• If you have recently migrated from another platform, factor in the period needed for purchase history to accumulate.
• Verify the click rate and purchase rate using the two standard reports before implementing the app.
• Identify the specific reason for needing an app among "configuration," "personalization," "A/B testing," and "measuring sales revenue."
• Calculate which is more advantageous between order-based and fixed-fee charging based on your monthly order volume.
• Confirm that the app you are considering does not have Plus limitations (LimeSpot Turbo) or Plus exclusive features (Rebuy Checkout Extensions).
• If Japanese UI is required, be prepared to choose between two options.
• Confirm that your theme is compatible with complementary blocks.

Summary

Four criteria are used for decision-making.

1. Related products are displayed by the native feature. First, confirm whether the display is by the native feature or an app. Many requirements can be resolved by setting up the free Search & Discovery with manual and complementary items.
2. The native limit is 10 items, 2 values, and manual settings. Knowing these three points allows you to immediately determine whether the request is within the app's domain.
3. Japanese stores have one less standard algorithm. For new products or recent migrations, consider using manual settings and collections to make up for the one less item.
4. Four reasons justify paying a monthly fee for an app. Configuration, personalization, A/B testing, and measuring sales revenue. If none of these apply, the native feature is sufficient.

The next step is to check the current click rate and purchase rate by opening the "Product recommendation conversions over time" report in the Shopify admin under Analytics → Reports. Without a baseline, you cannot evaluate the app's effectiveness.

The pricing, ratings, and official names of the applications can change. The values in this article are based on direct confirmation from the official documentation, the Shopify App Store, and the developer's official pages as of September 1, 2026. Please check the latest information before implementation.