Need help with product recommendations in Shopify PHP development
I’m working on building a Shopify application using PHP and I’m stuck on something. I need to fetch products that are related to what customers are viewing or buying - you know, like the “customers also bought” or “you might also like” kind of products.
I’ve been going through different PHP libraries and API documentation but I can’t seem to find clear endpoints or methods for retrieving these recommendation products. Most of the standard APIs seem to focus on basic product CRUD operations.
Has anyone worked with similar functionality before? I’m wondering if there’s a specific API endpoint I’m missing or if I need to build this logic myself by analyzing purchase patterns and product tags.
Any guidance on the best approach would be really helpful. Thanks!
hey! yeah, shopify doesnt have a built-in recommendation api, which sucks. most devs build their own using product tags or purchase history from the orders api. you could also try third-party tools like algolia or recombee if u want something ready-made.
the metafields approach works great, but product variants can be solid for cross-sells too. I’ve had good luck matching by vendor + similar pricing - it’s surprisingly effective for basic recommendations without needing complex algorithms.
Had this exact problem building recommendation features for several Shopify stores. Here’s what worked: I combined customer segments with product association analysis using the REST API. Don’t just look at purchase history - pull customer data and segment buyers by demographics and behavior, then map product relationships based on those segments. Premium customers cross-sell differently than budget shoppers, even for the same products. I used the Customers API with Orders to build segments, then stored recommendation mappings in a separate database table that updates weekly. This segmentation approach crushed basic ‘frequently bought together’ logic because it considers customer intent and spending habits. Performance was solid since the heavy processing runs offline - you just query your recommendation table during page loads.
Built cross-sell functionality for a Shopify client about six months back - went with a hybrid approach that worked pretty well. Started simple by pulling related products from shared collections and product types via REST Admin API, then added custom logic using cart abandonment and order history data. The trick was weighted scoring based on what customers actually bought together, similar price points, and complementary categories. Switched to GraphQL API for the product relationship stuff since REST was crawling with bigger catalogs. Best move was storing recommendation data in metafields and updating periodically instead of calculating everything live. Cut way down on rate limiting headaches and page loads got much faster. Takes some upfront work but barely any maintenance compared to running recommendation logic on every single page request.
Manual coding sucks when you need recommendations that actually convert.
Automation beats writing PHP logic for every scenario. Pull your Shopify data, run smart filters on purchase patterns, and auto-update product relationships without rebuilding.
Set up workflows that watch customer behavior live. Someone adds to cart? System instantly finds similar items by tags, price, or what others bought together. Updates happen automatically via API.
No more manual metafield updates or rate limit headaches. Set rules once, let it run.
Great for seasonal trends or new launches where manual tagging can’t keep up. System learns from real sales data and adjusts recommendations on its own.
Saves massive dev time vs building recommendation engines from scratch.
Check out Latenode: https://latenode.com
Had this exact problem last year building a custom Shopify solution. Manual approach gets messy fast with complex recommendation logic.
Automated workflow worked way better - pulls product data from Shopify, analyzes purchase patterns, and updates recommendations without touching PHP code.
Built a system that monitors customer behavior, processes orders, and auto-updates product recommendations based on tags, categories, and purchase history. Runs in background and feeds results back through webhooks.
Cool part is setting up rules like “customer buys product A, suggest similar tagged products” or “show top sellers from same category” without writing complex algorithms.
Saved me weeks of coding and handles edge cases automatically. Scales better than processing everything through PHP requests.
Check out Latenode for this: https://latenode.com
Shopify doesn’t have built-in recommendation APIs, so you’ll have to build this yourself. I’ve done this before - GraphQL Admin API beats REST hands down for product relationships since you can grab exactly what you need with fewer calls. Best approach I used was building a recommendation engine that digs through order data to find products people buy together, then saves those relationships in product metafields for fast access. You can also do basic recommendations by matching similar tags, categories, or price ranges. Cache your calculations instead of running them on every page load or you’ll hit rate limits fast. Takes work upfront but way better than real-time processing.