Personyze offers a rich suite of product recommendation algorithms, each crafted for specific pages and customer contexts across your site, emails, or app. These algorithms maximize relevance, increase order values, and boost conversions by combining individual visitor behavior with crowd insights. On any given page, you can display multiple algorithms within the same widget or across multiple widgets. You have full control over placement and display order, and you can even combine embedded on-page recommendation widgets with floating popups—for example, an exit-intent popup prompting visitors to enter their email to receive updates when a product goes on sale or is back in stock.

Here are examples of some of the algorithms available:

Homepage

Smart Homepage Recommendations

Show the most relevant products on the homepage for both anonymous and known visitors. For anonymous users, the engine leverages signals like location, weather, referral source, ABM data, and crowd activity to recommend items they’re most likely to buy. For known visitors, recommendations are personalized based on past transactions and preferences—creating a truly dynamic entry experience.

Most Popular Products

Promote bestsellers across the site or within categories relevant to the visitor, adding social proof and discovery value.

Recommended for You

Machine learning analyzes visitor behavior and crowd data to suggest the products each shopper is most likely to buy.

Product Page

Frequently Bought Together

Cross-sell products often purchased alongside the item being viewed.

Those Who Viewed This Also Viewed/Bought

Display alternatives or complementary items based on crowd behavior, keeping visitors engaged and helping them find the right product faster.

Those Who Bought from This Category Also Bought

Highlight items from other categories frequently purchased by shoppers who bought from the same category.

Cart & Checkout

Cross-sale Cart Recommendation

On the cart page, suggest items frequently purchased together with products already in the cart.

In Your Cart / Wishlist

Keep visitors focused on their shopping journey by displaying their current cart contents or saved wishlist items for convenient navigation and reminders.

Returning Visitors

View It Again / Buy It Again

Show products a visitor has recently viewed or purchased, making it easy to pick up where they left off or reorder essentials.

Promotions & Urgency

New in Stock

Promote newly added products, customizable by recency (e.g., last 7 days, since last visit) or by visitor interests.

Recently Discounted

Feature products that just went on sale, prioritized by the shopper’s interest profile.

Email Recommendation Algorithms

Now on Sale from Past Interests / Left in Cart

Shows items that recently dropped in price, based on what the visitor previously interacted with or left in their cart.

Cart Abandonment Drip

Reminds visitors of items left in their cart and suggests similar products, sent at intervals you define.

Time to Resupply

Prompts the visitor to re-order perishable or consumable items, sent on a recurring schedule of your choice.

Most Recommended for You

Uses machine learning and crowd patterns to suggest items the visitor is most likely to be interested in.

Cross-sell Based on Last Purchase

Recommends items that others bought after the same purchase, or products that have a cross-sale association in your catalog.

Newly Listed Items / Accessories for You

Highlights newly added products that complement items previously purchased.

Send My Cart to Email Button

Lets visitors opt-in to receive their cart by email as a reminder. If their email isn’t known, it can trigger an email capture popup, which can then be used for future remarketing campaigns.

Further Fine-Tune Recommendations with Filters

• Discount thresholds (greater or less than X%)
• Product feed attributes (interest, color, brand, price, etc.)
• Location-based rules (e.g., within X miles of a distribution center = free shipping)
• Visitor demographics (gender, age, client type, industry, etc.)
• Stock level thresholds (e.g., show items with fewer than X units left)

Product recommendation widgets are simple HTML, giving you the freedom to use any design you like. Personyze also provides a library of responsive templates—horizontal and vertical by default—with a visual editor to customize every detail to your brand style. Templates can optionally inherit your site’s CSS to speed up styling.

Templates include a wide range of options such as lazy-loading images, rotating multiple images on hover, displaying any product attributes from your catalog, and adding dynamic badges with logic, for example:

• New badge if the item was added in the past X days
• Recently discounted badge if the price dropped within X days
• Low stock badge if inventory falls below X units

Additional features include quick “add to cart” buttons so visitors can instantly add recommended items, save-to-wishlist directly from recommendations, multi-tab recommendation widgets, hover info cards, and more.

You also have full control over display volume—for example, setting the total number of recommended items shown, and separately configuring how many items appear per slide in a carousel or slider.

All recommendations are fully responsive and can be displayed as embedded blocks, carousels, or popups. They can also be combined with lead forms (e.g., showing cart items and asking for an email to save the cart) or countdown timers for special sales. Recommendations work seamlessly across sites, emails, and apps—using any HTML/CSS needed to achieve the exact look and feel.

Personyze allows you to upload past transaction data from offline purchases, physical stores, or third-party systems via file upload, feed synchronization, SFTP, or API. This enriches the recommendation engine and ensures that all available data is fully leveraged for more accurate and effective recommendations.

Uploading historical transactions can also jump-start your recommendation engine when you first sign up. For example, algorithms like cross-sell (“Those who bought this also bought”) normally require many live visitors to generate patterns. By importing past transaction data, you give the engine a rich dataset to work with immediately—delivering meaningful cross-sell and upsell recommendations from day one.

Every custom field you create—such as user location, industry, account type, or any other attribute—becomes a personalization tag that can be inserted directly into the recommendations widget or JSON recommendation set.

This allows you to dynamically tailor messaging, for example:

• Include the customer’s name in the title of their recommendations
• Display “New items from X brand” based on the last brand they viewed
• Show “Recommended for you because you purchased X in the past”
• Highlight “Most popular in City X today” based on the visitor’s location