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.
Where a traditional A/B test runs one experiment with one winner, an Agentic Campaign runs hundreds of simultaneous personalisations, each one tuned to a different audience and moment, and re-tuned every night as new evidence comes in.
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;
How the agent learns, what the control variant does, and what the reporting shows
A step-by-step guide to getting your campaign live
The available agentic strategies, and recommended set-up
How to interpret agentic campaign results
The problem this solves
Most eCommerce teams already know what experiences they want to run; email capture, social proof, shipping nudges, discounts. The hard part is deciding who should see each one, and when.
What you usually end up doing is picking a segment based on a hypothesis ("high-intent shoppers should see the discount"), launching it, watching results, tweaking the threshold, and layering on extra rules. It's trial-and-error, and it scales linearly with effort - more permutations mean more hours, more people, or both.
The hardest part is the trigger, not the treatment. Even if you agree that, say, struggling high-intent shoppers need help, you still have to decide when to intervene. Does the best moment differ by device, channel, landing page, or product affinity? Should you target the broad group, or only those with an extra qualifier? There's no obvious right answer - you have to test your way there.
This is especially acute on discounting. Set a blanket rule and you give discount away to shoppers who'd have bought anyway. Test your way to a crude intent threshold and you still haven't answered the question that actually matters commercially: when is the optimal moment to offer a discount to this specific shopper, in this specific session?
An Agentic Campaign treats that decision as the thing to optimise. Instead of testing treatments against pre-built segments, it auto-discovers the audiences and moments that matter, and keeps learning.
What this means for you
If you're on the experience team and feeling stretched
If you're on the experience team and feeling stretched
"I know we could be doing more, but I don't have the time."
You add your experiences, pick a strategy, and the agent handles segmentation, allocation, and timing. Most campaigns go from strategy choice to live in under an hour, and once they're running the ongoing maintenance is about five minutes a week to check the Performance Snapshot.
Common questions from teams like yours:
How much setup is involved? Minimal — pick a strategy, add experiences, go.
Will it conflict with campaigns I've already built? No, but you'll need to set up exclusions like you do with custom campaigns.
Do I need to understand the data science? No, but the full transparency is there in the AI Decision Log if you want it.
If you're leading digital or eCommerce and under commercial pressure
If you're leading digital or eCommerce and under commercial pressure
"The board wants 20% growth but my budget is flat."
Agentic Campaigns replace manual experimentation with continuous, always-on optimisation. You get uplift vs control measured the way you already measure it — but across hundreds of audience contexts simultaneously, not one test after another.
Common questions from leaders in your position:
What ROI metrics will I see? Uplift vs control, probability to beat control, and an annualised revenue projection — all in the Performance Snapshot.
Can I show this to the board? Yes. The dashboard is built for it — uplift, projections, allocation trends, AI Decision Log.
Does this replace my A/B testing tool? For continuous, always-on optimisation, yes. For one-off structural tests (e.g. a hero CTA copy decision), A/B testing still fits better. Many teams run both.
If you're in trading or commercial and fighting margin erosion
If you're in trading or commercial and fighting margin erosion
"We're addicted to discounting. Every sale costs us margin."
Set up multiple discount levels as separate experiences — 0%, 5%, 10%, 15%. The agent figures out who actually needs the nudge and who'd have bought anyway. With Revenue Per User as the goal, it's optimising for net revenue rather than just conversions, so discount cost is always factored in.
Common questions from teams managing promotions:
Can I still set rules? Yes — global matching criteria still apply. "Never show a discount in checkout" works exactly as you'd expect.
How does it handle peak trading? The rolling 90-day training window lets the agent adapt across trading conditions, and higher peak traffic typically speeds up learning rather than disrupting it.
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 |