Understanding Personalisation Signals in Liva 7
How Liva 7 reads shopper behaviour and product relationships to deliver smarter recommendations
Overview
Personalisation signals are the heartbeat of Liva 7's recommendation engine. They allow the system to interpret what shoppers are doing, what they're interested in, and how products relate to each other - all in real time.
For merchants, understanding these signals is key to unlocking better performance, smarter targeting, and more trust in the engine's decisions.
Behaviour signals: decoding shopper intent
These signals track what shoppers do on your site and help Liva 7 respond instantly with relevant content.
- Browsing behaviour
Every page view, product click, and interaction builds a profile of what the shopper is exploring. For example, if someone views three different hiking boots, Liva 7 will prioritise similar outdoor footwear in recommendations. - Cart behaviour
Adding or removing items from the cart is a strong signal of purchase intent. If a shopper adds a camera to their cart, Liva 7 might suggest compatible accessories like tripods or memory cards. - Purchase history
Past orders help personalise future visits. If a customer previously bought skincare products, Liva 7 may highlight new arrivals or complementary items in that category when they return.
These signals are active for every visitor, even anonymous ones, and update in real time as the session unfolds.
Behaviour signals are strongest when shoppers interact with multiple products in a short time. For example, viewing three different backpacks in one session tells Liva 7 to prioritise outdoor gear in recommendations.
Similarity signals - mapping product relationships
Similarity signals help Liva 7 understand how products relate to each other, even without manual tagging or bundling.
- Shared attributes - Products with the same tags, collections, vendors, or types are considered related.
- Co-purchase patterns - Liva 7 learns from your order history to identify items frequently bought together.
- AI vector analysis - Uses machine learning to detect deeper similarities in product descriptions, titles, and metadata.
Similarity signals allow Liva 7 to recommend a matching lamp for a sofa, even if they're in different collections - as long as they share design traits or are often bought together.
How signals combine in real time
Liva 7 doesn't rely on a single signal. It blends them to create a dynamic, session-aware experience:
- A shopper views three different backpacks - browsing signal activates
- They click on a specific hiking backpack - click signal increases its weight
- They add it to cart - cart signal triggers accessory suggestions
- They return a week later - purchase history and session memory personalise the homepage
Liva 7 recalculates recommendations every time a shopper takes an action - even within the same session - so blocks stay relevant as behaviour evolves.
Real-world examples for merchants
- Fashion boutique - A shopper browses three floral dresses. Liva 7 shows similar styles in 'Inspired by Your Browsing', then recommends matching sandals and bags in 'Complete the Look'.
- Electronics store - A customer adds a gaming laptop to their cart. Liva 7 suggests a mouse, cooling pad, and monitor in the cart block using co-purchase and cart signals.
- Home decor brand - A returning customer who previously bought a velvet sofa sees matching cushions and throws on the homepage, powered by purchase history and similarity analysis.
- Beauty retailer - A new visitor clicks on a skincare product from a Facebook ad. Liva 7 uses the referrer signal to prioritise trending skincare items and shows a 'Recently Viewed' block to reinforce interest.
These examples show how signals translate into real merchandising outcomes - helping you increase AOV, reduce bounce, and personalise without manual rules.
Why this matters for your strategy
Understanding personalisation signals helps you:
- Diagnose unexpected results - if a block shows surprising products, check which signals are active
- Design smarter strategies - use behaviour-heavy strategies on high-intent pages (like cart), and similarity-based ones on discovery pages (like homepage)
- Build trust in automation - knowing how the engine thinks makes it easier to let go of manual merchandising
- Guide the engine - use filters, exclusions, and targeting to shape how signals are interpreted
You don't need to manage signals directly - but knowing how they work helps you make smarter decisions about block placement, strategy choice, and targeting.
It's powerful, it's effortless
Personalisation signals are simply the behaviours shoppers show you as they browse. Liva 7 reads those signals instantly and adjusts the storefront to match what each person is interested in. Every view, click, scroll or cart action strengthens the engine’s understanding of what matters to that shopper in that moment.
You don’t need to guess what to promote or build manual bundles. The engine handles that by recognising what the shopper is drawn to and which products naturally fit together. The result is a storefront that feels intentional, relevant and driven by real behaviour rather than guesswork.
When every product shown is backed by genuine interest and smart product logic, personalisation stops feeling like automation and starts feeling like a better way to merchandise your store.
