UNRESOLVED
When Feedback Has To Matter
Paul Tucker leads support and success for ChiroCat with a team of five people he hired specifically to speak his customers’ professional language. He walks through the customer appetite thinking that shapes how he deploys AI, and the feedback problem he’s still working through.
Paul Tucker doesn’t have a chiropractic background, which is why he hires insurance billers who’ve worked inside chiropractic offices, people who can speak their lingo, front desk intake workflows, and secondary insurance billing in a way that earns credibility before the conversation is ten exchanges deep. The word on chiropractic social channels is that ChiroCat’s support is legendary. That reputation came from a hiring decision.
The feedback that stings
Paul described two failure modes he’s seen in CX teams when negative customer feedback arrives. The first is defensive dismissal, where the team absorbs the feedback, acknowledges a grain of truth, then shields itself by deciding the customer would have responded differently if they understood how hard the team was working. The second is over-internalization, the spiral where individual pieces of feedback start to feel like evidence that nothing the team does will hold.
What Paul’s after sits between those two. He wants feedback to land with enough weight to generate action, without landing so hard it distorts what the team believes it’s capable of. He described a specific case: a ChiroCat customer whose expectations the product genuinely couldn’t meet. The mismatch was in what the product could provide. His guidance to his team: let that sting just a little. Acknowledge, with some humility, that another product served that person better, and carry that forward without being destabilized by it.
He’s building tools to get upstream of this. Last quarter he loaded ChiroCat’s full customer conversation history into Claude and asked it to map the product’s friction points by theme. What came back shifted the company’s roadmap for the quarter, moving from a feature-first approach to a workflow-first one. He’s also built churn-tracking tools to flag leading signals before they become cancellation requests. The tools surface the signal. What Paul hasn’t figured out is what to do with that signal in a way that’s consistent, responsive, and not subject to the two failure modes he described.
You can’t automate trust, and you can’t automate relationships.
What Paul’s team built at ChiroCat
ChiroCat is a practice management platform for chiropractors and their teams, a market Paul described as neglected by software developers because it doesn’t generate high-ticket insurance billing at scale. He leads a team of five direct reports covering support, success, and onboarding, and he’s actively hiring to expand. The company uses Help Scout for customer conversations, Slack for internal communication, and a set of Claude-powered integrations he’s built himself. His description of the approach: “MCP all the things.”
The hiring philosophy is the most deliberate structural choice Paul’s made at ChiroCat. He knew early that he couldn’t train generalist support staff to speak chiropractic billing fluently. Chiropractic billing involves years of technical expertise: insurance submission, secondary billing to Medicare, and front desk intake workflows. When a chiropractor contacts ChiroCat’s team and the person responding already knows what that workflow looks like, trust is established quickly. Paul’s team earns it before they’ve had to prove anything else.
Brett described a parallel at Frame.io, where he hired video production specialists rather than generalist support staff, people who’d assembled professional cameras on set and worked across Premiere, DaVinci Resolve, and Avid. That background let the team meet customers where their actual workflow was, not just where Frame.io fit into it. A customer struggling with a Premiere sequence was often describing a workflow gap rather than a feature request, and someone with production experience could address that directly. Both teams arrived at the same conclusion from different industries: you can teach someone the support playbook. Teaching them the professional context that makes a customer feel understood is a different problem entirely.
On customer appetite, Paul’s position is that most companies measure AI deployment by deflection rate without accounting for whether deflection is what the customer actually wanted. His customers are chiropractors whose primary work is physical and relational. Their appetite for navigating an AI chat to resolve a billing question is different from that of a software operator, and building toward high deflection numbers without understanding that appetite is a retention risk that won’t show up in any dashboard until it’s already cost something.
He also drew a specific line about where AI belongs in his operation. His Claude-powered tools are internal: a churn tracker that flags at-risk accounts and manages remediation as a project, a brand voice tool that lets his support team draft responses in ChiroCat’s tone, and a custom Slack integration that reconciles to-do items across his messages and meetings. AI is doing work inside the operation, behind the customer interactions rather than inside them.
Paul Tucker’s hiring decision at ChiroCat is a specific answer to a question most CX leaders don’t ask directly: what’s the professional knowledge your customers assume the person helping them already has? In chiropractic practice management, that knowledge is medical billing and clinical administration. The tooling Paul has built inside his operation is sophisticated, and he uses AI heavily. The customer-facing side stays human because his customers’ appetite for AI support is low, and he’s learned to treat that appetite as a real variable rather than something to optimize around later.
Paul’s still working through the feedback problem. He has tools that surface signal from customer conversations at scale. What he’s still figuring out is the internal practice: sitting with what customers say, letting it matter enough to generate action, without letting it become something that slows the team or distorts its judgment. If you’ve built a process for that, or you’re closer to one than he is, he’s worth reaching out to.
That’s what Unresolved is for.
Links mentioned
Follow the show
Have a story to tell?
So many CX leaders are testing with AI and rebuilding with AI as their foundation. Every one of us have to reinvent everything that we’ve known about customer experience systems and processes.
You’re not alone, so let’s do this journey together.