Leverage pre-configured collaborative filtering algorithms, such as Alternating Least Squares (ALS) and Bayesian Personalized Ranking (BPR), to process and analyze your shopper signal data which then produces contextualized item and user specific recommendations. These algorithms come with templated training jobs and standardized pipelines, as well as highly customizable features, to serve a variety of use cases and help teams get the most out of recommendations.
No signals? No problem! Deep-learning vector-based recommenders are pre-configured. This recommender functions by looking at your product title names, descriptions and any additional metadata available. This job creates a model that places all products in the catalog into a vector space and creates item-for-item recommendations based on what products are near one another. This is a powerful tool to complement the signals-based recommenders and is a performant cold start solution when signals are not readily available.
Surface item-for-item recommendations from ALS, BPR and deep learning-based algorithms on your PDPs. Fusion Recommendations make it easy to post-process item recommendations for specific situations, such as only displaying items within a certain price threshold of the initial item.
Easily experiment with a variety of algorithm parameters and pipeline configurations. Leverage the experiments framework to assess algorithm performance and fine tune recommendations to your use case.
Cover the homepage with different types of personalized recommendations for your users. The ALS and BPR algorithms output contextualized items-for-user recommenders. You can surface those items on your home page in a “Recommended for You” zone, as well as in a variety of other locations throughout your website.
Leverage Fusion’s pre-configured trending algorithm to power a “Trending Now” zone anywhere on your site. The algorithm conducts statistical analysis on recent shopper behavior to uncover trending queries and items to be displayed in the zone.
Basic aggregation jobs and templates are available to power “Popular Items” and “Popular Queries” zones. The output from these aggregation jobs and templates are easily integrated into query pipelines so they can be placed anywhere on the website.
Guide users with higher converting and higher engagement queries. This recommender can be placed on a home page for a personalized effect such as “Search terms that may interest you” or on search result pages to the tune of “People also searched for.” This similar query method comes standardized with the same templates and pipelines as the other recommenders.
Research from Gartner reveals the ROI of personalization.
Las Vegas travel search engine Vegas.com increased page views and engagement by 63% . Bounce rates for the mobile site dropped 8%. Conversion rates from search to a reservation or ticket increased by 33%. The team has since built the desktop search experience with Fusion as well. Full case study…
The American sportswear and footwear retailer put Lucidworks to work and saw a 10% lift in add-to-cart, improved personalization, and better analytics to drive merchandising decisions for promotional events.
Fusion AI learns from your customers to predict the next best experience, accelerate the path to purchase, and reduce operational burdens. Learn more
Retrieve relevant results to challenging queries with Semantic Vector Search. Learn more
Merchandise and curate digital experiences with Predictive Merchandiser. Learn more