Great experiences start with the right recommendation. Whether it's a product, article, or offer, intelligent recommendations help customers quickly discover what matters to them — driving deeper engagement and measurable business impact.
Behind every effective recommendation is a balance of AI-powered personalization, real-time signals, and contextual understanding. Blueshift helps teams deliver personalized suggestions without complex setup or long development cycles.
By unifying behavioral data, catalog insights, and predictive models, Blueshift surfaces relevant content across key channels — email, push, SMS, in-app, and web — adapting as user behavior evolves.
Real results from leading brands
🛏️ Tuft & Needle increased email revenue by 181%.
🚗 CarParts.com boosted engagement by 400%.
These capabilities come together in what we call recommendation recipes — the foundation of Blueshift's personalization strategy.
Why Blueshift recipes drive results
Rapid deployment: Access 100+ AI-powered recipes mapped to industry goals and lifecycle stages.
Enterprise flexibility: Combine Blueshift's models with your own data science by importing external feeds.
Unlock the power of 100+ AI-powered recipes
⚡ Proven strategies across industries and lifecycle stages.
✨ Works across sessions, channels, and lifecycle stages.
AI-powered recipes for lasting customer impact — browse by vertical
Retail
Recipes for e-commerce and retail platforms, focused on product discovery and cart recovery.
Abandoned cart
Remind users and help them continue where they left off
Abandoned cart
Remind users and help them continue where they left off
Use cases
Remind users and help them continue where they left off.
Combine with similar items product recommendations to let them explore and discover more.
Use multi touch recommendations for high intent users and prevent lost revenue.
Requirements
Add_to_Cart event or equivalent with product_ids.
Purchase event or equivalent with product_ids.
Item catalogs with those product_ids.
Inputs
Timeline on add_to_cart events.
Timeline on purchase events.
How it works
Replay Items added to cart but not purchased during checkout process and bring back users to where they left off.
Brand affinity items
Popular, trending or recent items in affinity brands
Brand affinity items
Popular, trending or recent items in affinity brands
Use cases
Daily/Weekly newsletter with top/trending items in user affinity brands.
App/Site live content with top/trending items in user affinity brands.
Requirements
Item catalogs with brand.
Items activity/event timeline.
Inputs
Number of items to recommend per affinity brand (Default 2).
Brand items ranking to fetch candidates.
Site level backfill ranking (if recommended products less than return count). Set 'None' if not required.
How it works
Popular, Trending, Recent or Deal Items in user affinity brands.
Wishlist reminders
Wishlist reminder for items not purchased
Wishlist reminders
Wishlist reminder for items not purchased
Use cases
Wish-list reminders to users for items they wish-listed but did not purchase.
Requirements
Users activity with wish-lists and purchase events.
Inputs
User wish-list event and timeline.
User purchase event timeline.
How it works
Items added to wish-list but not purchased yet.
Derived event - price drop
Price drop reminders
Derived event - price drop
Price drop reminders
Use cases
Price drop reminders.
Combine with similar items to help users compare various products and let them explore.
Weekly / Monthly newsletters.
Requirements
Events saved on timeline with product_ids in payload.
Price change threshold.
Item Catalogs matching those product_ids.
How it works
Items with price change which have been browsed and not yet purchased by users.
Media
Recipes for streaming, publishing and content discovery, focused on engagement and personalized feeds.
Streaming - similar shows to recent activity
Daily/Weekly newsletter to users based on recent activity
Streaming - similar shows to recent activity
Daily/Weekly newsletter to users based on recent activity
Use cases
Daily/Weekly newsletter to users based on recent activity.
Suggested content to view based on past activity in app/email.
Requirements
Users recent activity timeline.
Item catalogs matching the products engaged (episodes).
Episode and Show catalogs linked through parent sku 3. Tags and title in the catalogs at episode and show level.
Inputs
User recent activity events and timeline.
Similar tag and similar title rankings to use.
Number of shows from recent activity to consider for similarity (default 1).
Number of similar shows to be recommended for each recent show engaged with (default: 2).
Site level backfill ranking (if recommended products less than return count). Set 'None' if not required.
How it works
Shows with similar title and tags to recently watched.
Feed type
Product-Product
Publishing - new items from authors you like
New items from authors based on prior purchases and wishlists
Publishing - new items from authors you like
New items from authors based on prior purchases and wishlists
Use cases
Alert users to new arrivals / launches from authors based on prior activity or wishlist.
Backfill other recommendations.
Requirements
Events with product_ids in payload.
Item Catalogs matching those product_ids.
Author attribute on items.
Inputs
Recent activity timeline.
Favorite Author Attribute on User profile 2. One of the precomputed rankings for each author.
How it works
New items from authors you purchased from or liked.
Tag affinity items
Popular or recent items from users affinity tags
Tag affinity items
Popular or recent items from users affinity tags
Use cases
Personalization related to tags inclination for popular / trending items.
Daily / Weekly newsletter feed to recommend popular / trending content by tags.
Requirements
Tags at sku and/or parent_sku level.
Users activity/event timeline.
Inputs
Number of items per tag affinity.
Popular Tags ranking to fetch candidate items.
Site level backfill (optional).
How it works
Popular or recent items related to the tags users have affinity towards.
Streaming - continue watching
Content users started watching but haven't finished
Streaming - continue watching
Content users started watching but haven't finished
Use cases
Remind users to continue watching video or audio or text content where they left off.
Combine with similar items product recommendations to let them discover related content.
Requirements
Media start event or equivalent with product_ids.
Media finish event or equivalent with product_ids.
Item catalogs with those product_ids.
Inputs
Timeline on media_start events.
Timeline on media_finish events.
How it works
Replay items watched on sites or apps but have not yet been completed and remind them to continue watching.
Marketplaces
Recipes for multi-seller platforms, focused on seller-based recommendations and time-sensitive listings.
New/trending from your favorite sellers
Re-engage users based on past activity
New/trending from your favorite sellers
Re-engage users based on past activity
Use cases
Re-engage users based on past activity.
Expand item discovery based on users activity.
Daily/Weekly customised feed to users.
Requirements
Item catalogs with seller.
User recent activity timeline.
Inputs
Recent event timeline.
Seller top items ranking to use for each seller.
Site level backfill ranking (if recommended products less than return count). Set 'None' if not required.
Apply size filter?
If yes, provide recent activity to use for determining size filter.
How it works
Based on recent activity sellers, recommend top ranking products from the sellers with size filter.
Ending soon items
Alert users to expiring items and promote related discounts
Ending soon items
Alert users to expiring items and promote related discounts
Use cases
Alert users to expiring items and promote related discounts.
Daily or Weekly Newsletters.
Backfill for other recommendations.
Requirements
Catalog of items.
end_date attribute in ISO8601 format.
Inputs
Site wide pre-ranked list of items sorted by end_date timestamp.
How it works
Items in catalog with end_date less than next few days indicating expiry or removal.
Next Best Items
Post activity follow up alerts
Next Best Items
Post activity follow up alerts
Use cases
Post activity follow up alerts.
Abandoned search/browse/cart follow ups.
Requirements
View or equivalent events with product_ids.
Item catalogs with matching product_ids and size in tags ( as applicable).
Inputs
Recent activity timeline.
Precomputed rankings based on other users with similar activity => Next best product, Cross-sell products.
Category backfill.
Site backfill.
How it works
Top items to consider after recent activity based on users with similar history and custom filters.
Feed type
Product-Product
Store name affinity items
Popular, Trending or Recent Items by user affinity store names
Store name affinity items
Popular, Trending or Recent Items by user affinity store names
Use cases
Daily/Weekly newsletter with top/trending items in user affinity store names.
App/Site live content with top/trending items in user affinity store names.
Requirements
Item catalogs with store name.
Items activity/event timeline.
Inputs
Number of items to recommend per affinity store name (Default 2).
Store Name Ranking.
Site-wide Backfill (if recommended products less than return count). Set 'None' if not required.
How it works
Popular, Trending or Recent Items by user affinity store names.
Automotive
Recipes for auto parts and vehicle-specific recommendations based on year, make and model.
Items by user vehicle's year, make and model
Items by user's vehicle
Items by user vehicle's year, make and model
Items by user's vehicle
Use cases
Items by user's vehicle.
Backfill Recommendations.
Daily Newsletters.
Requirements
Catalog with product_ids.
User year, make and model attributes.
Inputs
User Vehicle Year Attribute.
User Vehicle Make Attribute.
User Vehicle Model Attribute.
How it works
Items based on user vechicle's year, make and model.
Similar items to previous block by user's vehicle year, make, model and category
Best Sellers by user's year, make and model at a given category level
Similar items to previous block by user's vehicle year, make, model and category
Best Sellers by user's year, make and model at a given category level
Use cases
Best Sellers by user's year, make and model at a given category level.
Top viewed by user's year, make and model at a given category level.
Promotional content by user's year, make and model at a given category level.
Trending items by user's year, make and model at a given category level.
New additions by user's year, make and model at a given category level.
Homepage recommendations.
Requirements
Catalog with category and product_ids.
User year, make and model attributes.
Pre-computed rankings.
Inputs
Previous block name.
Item Category Level. The default value is Leaf level.
User Vehicle Year Attribute.
User Vehicle Make Attribute.
User Vehicle Model Attribute.
Vehicle year/make/model by Category Ranking 1. Select None, if not needed.
Vehicle year/make/model by Category Ranking 2. Select None, if not needed.
Vehicle year/make/model by Category Ranking 3. Select None, if not needed. 9 Vehicle year/make/model by Category Ranking 4. Select None, if not needed. 10. Site-wide Backfill. Select None, if not needed.
How it works
Top engaged items by user's vehicle year, make, model and previously interacted categories by previous block items.
Category affinity items by user's vehicle year, make, model
Best Sellers by user's year, make and model by category affinity
Category affinity items by user's vehicle year, make, model
Best Sellers by user's year, make and model by category affinity
Use cases
Best Sellers by user's year, make and model by category affinity.
Top viewed by user's year, make and model by category affinity.
Promotional content by user's year, make and model by category affinity.
New additions by user's year, make and model by category affinity.
Price ordered items by user's year, make and model by category affinity.
Homepage recommendations.
Requirements
Events with product_ids.
Catalog with brand and product_ids.
User year, make and model attributes.
Pre-computed rankings.
Inputs
User Vehicle Year Attribute.
User Vehicle Make Attribute.
User Vehicle Model Attribute.
Enter the number of user category affinities to consider.
Vehicle year/make/model by Category Ranking 1. Select None, if not needed.
Vehicle year/make/model by Category Ranking 2. Select None, if not needed.
Vehicle YMM Backfill Ranking. Select None, if not needed.
How it works
Top category affinity items by user's vehicle year, make and model.
Live content - Next Best Items to seed items + category level trending items
Next best items based on seed items in live context payload…
Live content - Next Best Items to seed items + category level trending items
Next best items based on seed items in live context payload…
Use cases
Live content on site or in app when users are actively in session to suggest next best items to purchase.
Requirements
Live Content request payload with seed_items.
Item Catalogs with matching seed_items.
Inputs
Name of event payload attribute for seed_items.
Name of event payload attribute for exclude_items.
Name of event payload attribute for exclude_categories.
Next best product ranking.
Category backfill level.
Upto 3 category rankings.
How it works
Based on seed items in recommendation context, recommend next best products and category top items. Category will be collected based on seed item and next best item. Category level can be defined as leaf, parent or root. If leaf - leaf level trending items will be added, if leaf + parent, then leaf + parent level trending items will be added. If leaf + parent + root, then the 3 level trending items will be added. Upto 3 category rankings can be selected.
Feed type
Product-Product
Online learning
Recipes for educational platforms, focused on course, teacher and content recommendations.
Courses for you
Homepage recommendations
Courses for you
Homepage recommendations
Use cases
Homepage recommendations.
Next Best Courses to take.
Requirements
User language attributes.
Catalog with relevant item ids.
Inputs
Use language filter. Tick checkbox, is teachers are to be filtered based on language.
User language attribute name. If selected, mention language attribute for user.
Catalog language attribute name. If selected, mention language attribute from catalog.
Courses Ranking.
How it works
Courses to consider based on user's language attributes at account level.
Live content - Popular items matching seed_tags in payload
Live content on site or in app when users are actively in session
Live content - Popular items matching seed_tags in payload
Live content on site or in app when users are actively in session
Use cases
Live content on site or in app when users are actively in session.
Requirements
Live Content request payload with seed_tags.
Item Catalogs with matching tags.
Inputs
Name of event payload attribute for seed_tags.
How it works
Based on seed tags in recommendation context, recommend other popular tag items.
Teachers for you
Post activity follow up alerts
Teachers for you
Post activity follow up alerts
Use cases
Post activity follow up alerts.
Incomplete search/browse/course follow ups.
Homepage recommendations.
Requirements
View or equivalent events with product_ids.
Item catalogs with matching product_ids.
Inputs
Recent activity timeline.
Use language filter. Tick checkbox, is teachers are to be filtered based on language.
User language attribute name. If selected, mention language attribute for user.
Catalog language attribute name. If selected, mention language attribute from catalog.
Precomputed rankings based on other users with similar activity.
Tags Similarity Ranking.
Site-wide backfill. For recommended teachers < return count. Select None, if not required.
How it works
Teachers to recommend based on other users with similar event activity and user's language attributes.
User feeds
Bring your existing user to item recommendations as feed imports
User feeds
Bring your existing user to item recommendations as feed imports
Use cases
Bring your existing user to item recommendations as feed imports.
Mix and match with pre-computed user rankings.
Alert users across channels including In-App or Onsite.
Newsletters.
User specific promotions / sponsored items.
Requirements
User profiles.
User feeds.
Inputs
User feed ranking.
Site-wide backfill (optional).
How it works
Personalized list of items per user as indicated in external feed uploaded by customer.
Feed type
User-Product
Location-based
Recipes that use location data such as city, state, zip code and coordinates.
City-state popular, trending, recent items
Popular or trending or recent items by city and state
City-state popular, trending, recent items
Popular or trending or recent items by city and state
Use cases
Show case your most popular or trending offerings to your new customers by specific city and state.
Alert all users to recent additions to your catalogs by specific city and state.
Alert all users to seasonal or sponsored promotions by specific city and state.
Win back your inactive customers with exclusive but popular items by specific city and state.
Backfill other recommendations.
Requirements
Events with product_ids in payload.
Item Catalogs matching those product_ids.
City and State attributes in item catalogs.
Inputs
City-State Name.
One of the precomputed rankings for each city state combination.
How it works
Popular, Trending, Recent items based on all user activity on your sites and apps by each city and state.
Live content - similar items to seed_items
Use products in seed items to recommend similar items
Live content - similar items to seed_items
Use products in seed items to recommend similar items
Use cases
Use products in seed items to recommend similar items.
Multi block personalization based on seed items.
Homepage recommendations.
Requirements
Events with product_ids.
Item catalogs with matching product_ids.
Inputs
Seed Items Array.
City-State Level Ranking.
Zipcode Backfill. Select None, if not required. For recommended items count < return count.
Site-wide backfill. Select None, if not required. For recommended items count < return count.
How it works
Items related to seed items with backfills. Similarity is based on user session and transaction activity with other users.
Items based on user lat/long
To provide recommendations for new users with insufficient history
Items based on user lat/long
To provide recommendations for new users with insufficient history
Use cases
To provide recommendations for new users with insufficient history.
Requirements
User profile with lat/long.
Product catalog with lat/long.
Inputs
User lat/long variables.
Radius for search.
How it works
Retrieve lat/long from user profile Find nearest products. Since this is meant for new users, there will be no filters as there will not be sufficient activity. For recent activity based recommendations, use other recipes.
Similar Items per alert attributes with custom filters per selection
Similar items per alert attributes set by user with custom filters…
Similar Items per alert attributes with custom filters per selection
Similar items per alert attributes set by user with custom filters…
Use cases
Use alert attributes to recommend items that fit that at regular cadence via newsletter Use alert attributes to recommend items that fit when user visits app/website.
Requirements
Events with product_ids.
Item catalogs with matching product_ids.
User profile with alert attribute.
Inputs
Previous block / recent activity.
Custom filters (bedroom, price, property type).
Supplement with user affinity categories?
Custom Priority.
How it works
Items on the basis of alert attributes supplemented with recent activity products that are provided as previous block. Similarity is based on user session and transaction activity of all users. Various filters to be applied can be selected.
Cross-industry
These recipes work across industries and use common personalization patterns.
Next best items/Similar items using LLM
Cross-catalog recommendations based on rich text similarity using LLMs
Next best items/Similar items using LLM
Cross-catalog recommendations based on rich text similarity using LLMs
Use cases
Cross-catalog recommendations based on rich text similarity using LLMs.
Requirements
Catalogs with rich text data.
Inputs
User recent activity / previous block.
Next best item/Similar Item LLM rankings.
Backfill ranking based on category/site if insufficient results.
How it works
Recommend items from other categories based on text similarity using LLMs.
Recent activity
Wishlist reminders
Recent activity
Wishlist reminders
Use cases
Wishlist reminders.
Re-order reminders.
Replay Purchase history.
Replay Cart history.
Replay Watch / Read / Stream history.
Requirements
Events saved on timeline with product_ids in payload.
Item Catalogs matching those product_ids.
How it works
Use items and their attributes from last 366 days of events to personalize templates and messaging across all channels.
Site-wide popular, trending or recent items
Popular, trending, recent items across all of your sites and apps
Site-wide popular, trending or recent items
Popular, trending, recent items across all of your sites and apps
Use cases
Show case your most popular or trending offerings to your new customers.
Alert all users to recent additions to your catalogs.
Alert all users to seasonal or sponsored promotions.
Win back your inactive customers with exclusive but popular items.
Backfill other recommendations.
Requirements
Events with product_ids in payload.
Item Catalogs matching those product_ids.
Inputs
One of the precomputed rankings across site wide activity or catalogs.
How it works
Popular, Trending, Recent items based on all user activity on your sites and apps.
Category affinity items
Daily or weekly newsletters
Category affinity items
Daily or weekly newsletters
Use cases
Daily or weekly newsletters.
Backfill for other recommendations.
Requirements
Events with product_ids.
Catalog of items with those product_ids.
Category attribute for each item in the catalog.
Inputs
Pre-computed category level rankings.
Number of items to consider per category. Default value is 2.
Site wide backfill if user does not have sufficient affinities or items. Select None, if not required.
How it works
AI generated list of items personalized to each user based on categories they have affinity towards. The list of categories is updated daily based on user activity across all sites and apps.
Choosing recipes by user lifecycle
Recipes can be mapped to different stages in the user lifecycle. Consider selecting a recipe based on your intended audience.
User lifecycle stage | Sample recipes |
New users |
|
Active users |
|
Intermittently active users |
|
Churned users |
|
All users |
|
What's next
Browse our library of ready-to-use recommendation recipes designed for common use cases across industries.
Explore recommendation recipes →