John Larson helped fund, shape, and scale product-led growth. As one of the most active investors in the SaaS ecosystem—and a Zipwhip co-founder during its rise and acquisition by Twilio—John was deep in the messaging and software trenches before most people knew what CPaaS stood for.

After Zipwhip, John could’ve ridden off into the sunset—and no one would’ve faulted him. But he was still looking for problems to solve. And he found them at Ambassador, where he is now Chief Strategy Officer.

There, he helped launch HIRO, an agentic AI designed to listen, respond, and act on what customers are actually saying. Not just to improve NPS. Not just to drive referrals. But to orchestrate action in real time—across data systems, teams, and even other AI agents.

Because in today’s world, customers don’t just want automation. They want orchestration.

And in this edition of One Expert, One Topic, John talks about conversation-led growth, the rise of agentic AI, and why the future isn’t single-use bots—it’s agents that listen, learn, and work together to drive outcomes.

About The Series

This is the thirty-first installment in the One Expert, One Topic series, where field experts break down one big idea using Matt Abrahams’ What / So What / Now What format. Written instead of recorded, so you can actually absorb it.

What

Product-led growth (PLG) helped SaaS companies scale by removing friction and letting the product do the selling. But in John’s view, we’ve already moved on. He calls the next wave conversation-led growth.

That means every feedback loop starts with a real conversation. It means agents—like HIRO—listen to what users say (or don’t say), measure tone, detect sentiment, and respond with clear next steps. “HIRO started by just listening carefully,” John says. “Because the most precious resource we have is our time.”

Today, that conversation might trigger a referral, a reward, or a support escalation. But tomorrow? The agent won’t just act on one insight—it will orchestrate outcomes across your tools, data, and AI stack.

Here’s how John draws the line:

• Agentic AI and orchestration layers represent different approaches to building and managing AI systems.
• Agentic AI focuses on creating autonomous, goal-driven agents that can act independently.
• Orchestration focuses on managing and coordinating the interactions between agents, systems, and humans within a larger ecosystem. It’s the difference between empowering one assistant—and getting all your assistants to work together.

Ambassador is building both. HIRO is an agent. But more importantly, it’s becoming the orchestration layer that pulls everything together: CRMs, loyalty engines, survey tools, social media channels, even other agents. The agent-to-agent architecture is the next big unlock—and HIRO is being designed to lead it.

So What

SaaS companies today are sitting on mountains of data. Most of it is stale or wrong. And most feedback systems—surveys, NPS, forms—fail to capture what people actually feel or need.

An example: One of the big three U.S. Carriers, an Ambassador customer, was sending out 100,000 surveys a month within a single business unit. But the data wasn’t trustworthy—users were just clicking through to get the incentive. To make sense of the results, this operator had to pay human agents to call people and dig deeper.

HIRO can replace that entire process with a real conversation. It captures sentiment, understands nuance, and acts on the result. Instead of chasing scores, Ambassador created the Ambassador Score—a next-gen metric based on interaction quality, emotional tone, and observable follow-through.

Ambassador replaces that with conversations that do something. Instead of collecting a score, HIRO understands context, follows up, measures tone, and drives outcomes in real time. The customer gets help. The brand gets a referral. A loyalty offer gets triggered. And the experience feels effortless—because the agent already knows what to do.

That’s orchestration.

John believes this shift mirrors the early days of SaaS: those who build the right infrastructure will win. It’s not just about launching an AI bot. It’s about connecting that bot to everything else—and letting it pull the signal from the noise.

As John puts it: “You’re not solving one thing. You’re stitching everything together.”

Now What

If you’re a SaaS company, brand, or mobile operator, here’s what John recommends:

1. Focus on orchestration, not just automation. Building a smart agent isn’t enough. You need the layer that talks to all your other tools—and to other agents.
2. Use AI to fix your data. Most enterprise data is full of holes. Conversations with agentic AI can validate, correct, and enrich it in real time.
3. Design for outcomes, not clicks. Don’t chase opt-ins you can’t convert. Don’t send messages that don’t respect the moment. “Never send a message unless it’s highly contextual,” John says.
4. Recognize that we’re all influencers now. HIRO doesn’t just capture the first referral—it tracks the second, third, and fourth layer of influence. “We’re finally able to measure what happens after the share and build a social commerce graph,” he says.

The end game? Ambassador wants to be the most connected AI feedback network on the planet. That means talking to internal agents (like Microsoft Copilot), external agents (like HIRO), and the open messaging ecosystem—including RCS. It means brands can finally stop blasting and start listening—while orchestrating meaningful actions behind the scenes.

Because the future of growth isn’t about sending more. It’s about listening better—and moving faster when you hear what matters.