Let AI do the heavy lifting

Relewise’s machine learning uses live session signals, a shopper’s history, and cohort trends to rank items. Our algorithms update rankings continuously, so you can cut manual merchandising while keeping control through guardrails you set.

Pages stay fast with one batched API request per view. The same profiles and signals power your site and app, keeping recommendations consistent.

A few recommendation types and typical placements

I wonder if this does anything?

"Personal Products"

"Purchased With"

"Viewed After Viewing"

"Popular Products"

"Personal Content"

"Similar Products"

Top picks for you - "PersonalProduct"

Personal recommendations based on the individual shopper’s profile and behavior, plus site trends. Raises relevance at entry points and deepens sessions.

Where to place it: Home and account pages; can also work as a secondary rail on PDP.

Frequently bought together - "PurchasedWith"

Suggests items often bought in the same order as the current item. Grows basket size with natural complements.

Where to place it: Product detail page and cart.

The crowd favorites - "PopularProducts"

Surface crowd favorites by category or across the site from recent activity. This recommendation highlights proven demand and drives more clicks.

Where to place it: Add to home and category pages.

Content that helps decide - "PersonalContent"

Recommends helpful content such as guides or articles alongside products, or products inside content. Answers questions that block purchase, which can reduce returns and increase confidence to buy.

A good placement: Product detail pages and editorial articles.

Smart spec-accurate swaps - "SimilarProducts"

This recommendation is especially useful in B2B or catalogs with detailed specifications, such as in the medical industry. It focuses entirely on product attributes to deliver accurate, attribute-based alternatives to the item being viewed.

Where to place it: Use on PDPs and PLPs to keep interest if the anchor item for any reason isn’t right.

Merchandisers stay in control

Stay in control by tuning recommendations to match your business goals. Boost or downplay items based on margin, brand, availability, or campaign tags, and pin or exclude products where needed. Stock, price, and freshness filters keep results relevant, while caps stop the same item from showing up in multiple carousels. Simple sliders let you balance algorithmic relevance with your priorities.

Performance and scale

Our API is built to stay fast, even with large catalogs. Multiple carousels can be batched into a single request per page view, with items automatically deduplicated across them. Pagination and cursor-based fetching handle big result sets, and compact JSON responses keep payloads light. The result is quick page loads, efficient requests, and recommendations that scale with your traffic.

Developer-friendly

Built on MACH principles, Relewise integrates with every e-commerce platform, personalizing product recommendations, product and content search, category listings, newsletters, etc. On top of that, our engine is lightning fast. We are confident we are one of the fastest engines out there, with an average server response time of less than 0.002 seconds! From the first to the last request, no caching is involved!

Important info!

Data & compliance

We're ISAE 3000 audited

Relewise is independently audited under ISAE 3000 on an ongoing basis. An external auditor regularly reviews the controls behind our AI features and verifies that customer data is processed securely and in line with GDPR, providing up-to-date third-party assurance for your team.

Frequently asked questions about Relewise Recommendations

Turning a blind eye to human rights violations erodes justice and equality, necessitating an urgent call to address and overcome these challenges.

How are recommendations personalized?

Relewise recommendations adapt to individual user behavior (views, searches, cart actions, purchases) combined with overall product popularity. They can also be segmented by classification (e.g., country, B2B role).

Can I control or override recommendations?

Yes. Merchandisers can shape recommendations through rules that let them boost, bury, pin, or filter products. This means business goals can be layered on top of the algorithm, for example by promoting high-margin products or removing items that are out of stock.

What data do we need to provide for recommendations?

To get recommendations on your site, Relewise needs a product catalog with IDs, categories, and attributes, as well as user behavior data covering product views, searches, cart activity, and orders. Historical order data can also be imported to give personalization an immediate head start.

Do recommendations work for B2B as well as B2C?

Yes. In B2B, recommendations can be tailored to roles, assortments per customer, or even very large catalogs, as seen with companies like Ahlsell and EET Group. The same system also adapts easily to B2C environments, giving both types of businesses the same personalization power.

How do recommendations impact business metrics?

The results are measurable. Personalized recommendations typically increase average order value by encouraging cross-selling and upselling, reduce drop-off by improving product discovery, and drive long-term growth through higher customer lifetime value.