Meet your new agent. With Agentic Campaigns, you choose the experience and define the strategy - the agent handles who sees what, and when. The result? Always-on optimisation that gets smarter over time.
Unlike Custom Campaigns, where you define the segment and timing yourself, an Agentic Campaign does that work for you. There’s no fixed end date, no manual segment-building, and no need to pick a winner.
Find out more about;
Agentic vs Custom: what’s the difference?
| Agentic | Custom |
Who defines the segment? | The agent | You |
Who decides timing? | The agent | You |
Campaign types available | Standard only | Standard, Dynamic, Sequenced |
How experiences are built | Pre-built templates, third party triggers, custom code | Pre-built templates, third party triggers, custom code |
What happens when an Agentic Campaign goes live?
When your campaign is eligible to run, the agent works through a series of steps before it begins optimising:
Data check; It will confirm there's enough recent data to learn from. If not, it will try again later.
Experience check; The agent will understand the content, behaviour and aim of the experiences you've added.
Audience grouping; It will apply your global rules and create the segments to receive your experience.
Timing setup (if enabled based on chosen strategy); Creates moment-based groups to optimise when experiences are shown, not just what is shown.
Learning begins; The campaign will go live, and the agent will begin comparing performance across experiences and visitor groups. Shifting traffic automatically toward what works best.
How does the agent learn?
It starts by sharing traffic fairly across your experiences so it can collect performance data
It measures what works best for each visitor group separately - one experience can win for one group and a different one for another
As results become clearer, it shifts the majority of traffic to those better performers while keeping some traffic on alternatives to continue learning
It uses shared learning. This means insights from one group can inform decisions in another, especially when data is sparse, to speed up optimisation
It tracks a confidence level: low confidence means broader testing; higher confidence means more focus on current best performers