How Smart Content works behind the scenes
Understand the mechanics of how Smart Content is delivered, optimised, and improved - so you can trust what’s happening and make smarter decisions.
Overview
Smart Content might look simple on the surface - you create a banner, assign a segment, and publish. But under the hood, Liva 7 is running a real-time personalisation engine that decides who sees what, when, and why.
This article explains how that engine works behind the scenes, so merchants can better understand what drives performance, how targeting decisions are made, and how to guide the system effectively.
Think of Smart Content as a living system. It doesn't just display content - it learns from every interaction and adapts automatically.
The 3 systems that power Smart Content
Smart Content is driven by three interconnected systems that work together to deliver and optimise content in real time:
1. Content delivery engine
This system controls when and where content appears on your storefront. It uses:
- Page type - Homepage, product page, collection page, cart, etc.
- Device type - Desktop or mobile
- Trigger - On page load, after scroll, or after delay
- Frequency - Show once per session, once per day, or always
These rules are set when you create content. They ensure your message appears in the right place at the right time - without slowing down your site.
2. Audience matching engine
This is the personalisation brain. It evaluates each visitor in real time using:
- Visitor type - New, returning, or known customer
- Behavioural signals - Products viewed, time on site, cart activity
- Purchase history - Past orders, frequency, value
- Segment membership - Based on your custom or AI-generated segments
The engine then selects the most relevant content block for that visitor from your available content library.
The more diverse your content and segments, the more options the AI has to personalise effectively.
3. Performance feedback loop
Every interaction is tracked and fed back into the system:
- Views, clicks, and conversions are logged for each content block
- AI Quality scores are updated based on how well content performs for its audience
- Targeting decisions are refined over time to prioritise high-performing content
This is why content often starts with a 'Creating a baseline' label - the AI is still learning how well it fits its audience.
What happens when you publish new content
When you launch a new Smart Content block:
- It's added to the delivery engine with its targeting and display rules.
- The AI begins testing it with a small portion of the intended audience.
- It monitors engagement - views, clicks, and conversions.
- If it performs well, it's shown more often. If not, it's deprioritised or flagged for review.
This process ensures that only the most relevant and effective content gets scaled across your storefront.
How the system handles multiple eligible content blocks
If more than one content block is eligible to appear in the same placement, the AI uses a ranking system based on:
- Relevance score - How well the content matches the visitor's profile
- Performance history - Past CTR and conversion rate
- Freshness - Newer content may be prioritised for testing
- Pinned status - Manually pinned content overrides AI ranking
You can pin content manually if you want to override the AI and guarantee visibility for a specific campaign or message.
A system that learns while you grow
Smart Content isn't just a tool for launching banners - it's a dynamic system that learns from your shoppers and adapts in real time. As a merchant, your role is to guide the system by:
- Creating content for different segments and goals
- Monitoring performance and AI Quality scores
- Iterating based on what the data tells you
- Trusting the AI to handle the complexity of targeting and delivery
Over time, this system becomes more accurate, more efficient, and more profitable - helping you scale personalisation without scaling your workload.
