The Retention Flywheel Is Real. We Helped Build One.
By Hannah McKernon, Head of Digital Marketing at Maybe*
Polly Barnfield OBE joins groa° and Octane AI at Innocos Events to talk zero-party data, DTC retention, and what it actually takes to turn a quiz into compounding revenue.
Most AI conversations end with a dashboard.
Ours ends with finished work.
That's why, when Emma Powell and Heidi Henneman at groa° asked Polly Barnfield, OBE, Maybe* CEO, to join them on stage at the Innocos Event in Paris, the answer was yes. Because this wasn't an event talking about AI potential. It was an opportunity to talk about AI proof.
What They're Talking About
14:40 - Agentic Zero-Party Data & The Retention Flywheel How Octane AI and groa° turned a DTC brand's quiz into a compounding revenue engine
Contributors:
Emma Powell, Founder & CEO, groa°
Heidi Henneman, Co-Founder & CMO, groa°
Aoife Teague, Director of Partnerships, Octane AI
Polly Barnfield OBE, Founder and CEO, Maybe*
Why This Panel Matters
There's a gap between AI that gets talked about and AI that gets work done.
The DTC world has felt that gap hard. Brands are sitting on rich zero-party data: quiz answers, preference signals, declared intent, and watching it disappear into a stack that doesn't connect. The result is a first-time buyer who never comes back. A VIP who gets the same message as someone who bought once eighteen months ago. Revenue that resets every cycle instead of compounding.
groa° built the Retention-First Growth® Flywheel to close that gap. Five connected orbits (Capture, Activation, Value Core, Loyalty, Reactivation) running autonomously, in real time, within clear profitability guardrails. Not campaigns. Architecture.
Octane AI brought the quiz data. groa° brought the system to govern it. The result: a compounding revenue engine, not a one-time conversion.
That's the kind of AI that interests Maybe*.
What Polly Will Bring to the Stage
Polly has spent the last two years inside 1,000+ businesses asking a single question: where does AI actually deliver finished work?
The answer is rarely where people expect.
It's not in the headline tools. It's not in the impressive demos. It's in the operational layer: the AI that sits inside real workflows, follows clear rules, and produces visible output that teams can act on.
The groa° and Octane AI story is a version of exactly that. Zero-party data from a quiz. An agentic OS that reads the signal and governs the next action. Revenue that doesn't reset. It compounds.
That's not AI theatre. That's AI that gets work done.
The Bigger Pattern
What groa° calls Replacement Growth, we see everywhere.
Brands, agencies, and businesses spending more each cycle to replace customers they already paid for. First-time buyers who don't return. Retention that depends on manual effort and hope.
The fix isn't another tool. It's an architecture that governs what happens next: automatically, at signal speed, within rules the business sets.
groa° built that for ecommerce. Maybe* builds that for every team that runs on Slack or Microsoft Teams.
The underlying truth is the same: AI stops being impressive and starts being valuable the moment it operates inside real work, reliably, without friction.
Come Find Us
If you're at Innocos Events, join the panel in the room at 14:40.
If you're not, follow the conversation.
And if you're sitting on zero-party data that isn't working hard enough yet, start with groa°.
Because 70% of first-time buyers never returning isn't a campaign problem.
It's an architecture problem.
And architecture is fixable.
Maybe* builds AI Agents that get work done, operating inside Slack and Microsoft Teams, connecting to the tools teams already use, completing real business work without friction.
