An extra member of the team: How Plungie scaled always-on customer conversations with Splash
Plungie makes something the building industry long thought impossible: a high-quality swimming pool that can be built in two days rather than six months. Its precast pool shells, protected by a global patent, remove one of the single biggest reasons home builders avoid pools - time. From its base in Queensland, Plungie is scaling that idea across Australia and the United States, selling not only to homeowners but to architects, designers, developers, home builders and a trusted network of installers it calls Plungie partners.
Scaling a genuinely new category is demanding, and Plungie does it with a relatively lean team.
“We lean heavily into tech, and where we can find value and still deliver a great customer experience, we will take full advantage of the capability.” says Nic Fry, who joined as Chief Marketing Officer in 2026 after nearly 25 years in global marketing.
The challenge: high volume, a complex category, juggling two very different markets
Every inquiry matters when you’re building a category. But pool projects are deceptively complex - access, siting and specification can each move the price significantly - so customers and builders arrive with detailed questions, often outside business hours, and across two markets that behave very differently. Handling that volume well, without a large support team, was the problem Plungie needed to solve. The goal, as Fry describes it, is to “find efficiencies and use tech to answer questions, triage, and solve issues, allowing our human team to focus on higher value solutions.”
The solution: Splash, designed to keep humans in the loop
Plungie turned to Aircall’s AI Voice Agent, aptly naming theirs Splash. From the outset the design was human-centric. Every visitor to Plungie’s website is given three clear choices: call the team, book a meeting, or talk with Splash - and it is always obvious which option connects them to AI.
“The human element is really important for us, and that always needs to be an option,” says Fry.
Rather than one catch-all system, Plungie runs a family of purpose-built agents: a website inquiry line, after-hours lines for both Australia and the US, a after-sale triage line, and an installation line that answers common questions on behalf of its team, before routing anything that genuinely needs them. Tone is tuned to each market, breezier in Australia, more professional in the US, based on what each audience responds to. Qualifying leads flow directly into HubSpot, ready for the sales team, through Aircall’s native integration.
Implementation: From pilot line to always-on layer
In fact, Plungie didn't switch everything on at once. The first Splash line was live and taking real calls quickly but the real work sat behind the scenes, in building and testing the knowledge base that tells Splash what to say. After-hours was the obvious place to start, giving customers a better experience for inquiries that land outside business hours. From there, Plungie added purpose-built lines over the following couple of months, refining the guardrails as it went, mapping the most common questions Splash could answer confidently, and making sure anything involving a frustrated customer was routed to a person early in the customer journey.
Much of that came together through Plungie’s partnership with Neighbourhood, both an Aircall and HubSpot partner, which scoped the objectives and worked alongside Aircall to deliver including the native HubSpot integration that feeds qualifying leads straight to the sales team. Michael Lance, Customer Experience Manager at Plungie is candid that while the earliest lines were a little clunky; the payoff came as the knowledge base matured and Splash learned from real conversations:
“It’s pleasing to see the learning evolve with consumer interaction.”
The results: Around-the-clock coverage that gets better with age
In its first six months, Splash has become a core part of how Plungie handles customer calls, fielding well over 1,600 conversations across inquiries, installation questions and after-sale support. The after-hours lines now cover nights, weekends and holidays in both markets without Plungie needing an on-call roster and on the US after-hours line, more than half of calls are resolved with no human involvement at all, in under 90 seconds on average.
Just as importantly, the agents improve as they learn. The Australian after-sale line’s goal-completion rate climbed from just under half to around 70% over a few months, even as volumes stayed high. And Plungie’s newest agent, the Australian website inquiry line, was the fastest out of the gate: within two months of going live it was resolving more than eight in ten calls on its own, with none needing a transfer.
“It’s really just an extra member of the team. It triages - like someone at a hospital emergency - so that when you do get to speak to a human, you can spend meaningful time with them. It’s not about cutting roles; it’s about using our people to keep scaling the business.”
-Nic Fry, Chief Marketing Officer, Plungie
That framing matters to Fry. Splash takes pressure off Plungie’s team, freeing them for the complex, high-value conversations where a human makes the difference, giving a better experience for both employees and customers alike.
Looking ahead
Plungie is already planning its next experiments: an AI Messaging Agent for WhatsApp and SMS, and outbound use cases such as inviting satisfied customers to leave reviews. Fry also sees Splash shaping the company’s wider content strategy where every question the agents can’t yet answer points to content worth building.
“I love the idea of having this as one of our content streams,” Fry says, “that helps us solve problems and deliver a better experience, but also helps inform where we invest.”
Nic Fry’s advice to marketers testing AI voice
Keep it human-centric. Always give customers a choice, and make it clear when they’re speaking to AI.
Start experimenting early. You learn fastest by launching, seeing what the agents can and can’t solve, then optimizing.
Treat your agents as a content stream — the gaps they surface show you exactly where to invest next.
See what AI Voice Agents can do for your support and sales teams: Try Aircall today