How the Personalisation Engine works
A clear, merchant‑friendly guide to how Liva 7 thinks, learns and adapts across your entire storefront
Overview
The Personalisation Engine is the intelligence layer that powers every recommendation, message and bundle inside Liva 7. It analyses shopper behaviour, product relationships and session activity to decide what to show, when to show it and who should see it.
This article explains how the engine works, how it chooses products and how its core settings shape the personalised experience across your store.
How the engine makes decisions
The engine analyses behaviour, context and product data to decide what each shopper sees.
The Personalisation Engine evaluates signals such as browsing patterns, product interactions, cart behaviour and session history. It combines these with product relationships and store‑wide rules to generate the most relevant output for each visitor. Every decision is made in real time, adapting as the shopper moves through your store.
The engine updates recommendations instantly as behaviour changes.
How Liva 7 chooses products
Product selection is based on relevance, intent and product relationships.
The engine looks at what the shopper is viewing, what similar shoppers have purchased and how products relate to each other. It prioritises items that match the shopper’s intent, using strategies such as affinity, complementary items, trending products and frequently bought together patterns.
The engine always aims to show the most relevant product at that moment.
Understanding personalisation signals
Signals are the behaviours the engine uses to understand shopper intent.
Signals include page views, product clicks, time on page, cart additions, search terms and more. Each signal helps the engine understand what the shopper is interested in and how strong that interest is. Signals vary by page type, so the engine interprets behaviour differently on product pages, collection pages and the cart.
Signals tell the engine what matters most to each shopper.
Engine Status and algorithm settings
Engine Status controls whether personalisation is active and which algorithm is used.
Merchants can toggle the engine on or off, choose the algorithm and set minimum and maximum recommendation counts. These settings define the overall behaviour of the engine and ensure recommendations match your store’s goals.
The engine must be turned on for widgets, content and bundles to personalise.
Session behaviour and time decay
Session settings control how long shopper behaviour remains influential.
Time‑based decay determines how quickly older actions lose importance. A shorter decay makes recommendations more reactive, while a longer decay helps the engine remember broader preferences. Anonymous personalisation can also be enabled to personalise for visitors who are not logged in.
Time decay balances short‑term actions with long‑term preferences.
How the engine works with rules
Rules shape what the engine is allowed to show.
Pinned products, excluded products, fallback behaviour and audience targeting all influence the engine’s decisions. Rules act as constraints or priorities, ensuring recommendations stay aligned with your merchandising strategy.
Rules guide the engine without replacing its intelligence.
How the engine powers widgets
Widgets use the engine to generate personalised product recommendations.
Each widget has its own strategy, page type and display settings. The engine decides which products to show inside the widget based on signals, rules and session behaviour. If multiple widgets exist for a page, the engine selects the most appropriate one.
Widgets are the visual output of the engine’s product decisions.
How the engine powers Smart Content
Smart Content uses the engine to decide who should see each message or banner.
Segments, triggers and frequency rules determine when content appears, but the engine evaluates behaviour to decide whether the shopper qualifies. This ensures messages feel relevant and well‑timed.
Smart Content personalises communication, not product recommendations.
How the engine powers Smart Bundles
Smart Bundles use the engine to build product combinations that increase AOV.
The engine identifies complementary items, affinity patterns and frequently bought together relationships. It uses these insights to generate bundles that feel natural and helpful to the shopper.
Bundles use the same intelligence as widgets but focus on combinations.
Widgets are the visual output of the engine’s product decisions.
How all engine‑powered systems work together
The engine unifies recommendations, messaging and bundles into one personalised experience.
Widgets recommend products, Smart Content communicates with shoppers and Smart Bundles increase order value. The engine ensures all three systems adapt to behaviour consistently, creating a seamless personalised journey across your storefront.
The engine is the single intelligence layer behind every personalised element.
