Episode Summary
In this episode of The Flywheel Effect, host Brent Sonnek-Schmelz and co-host Matt Bernath sit down with Jason Griffing, Director of Product at OneVision Resources, for a wide-ranging conversation that connects two of the most consequential forces reshaping the custom integration business: artificial intelligence and the economics of service. Jason opens with an inside look at how OneVision has woven AI through every layer of its product, engineering, sales, and support operations — and shares practical playbooks any integrator can adopt, starting tomorrow.
From there, the conversation pivots to the service challenge that has quietly defined the CI industry for decades. Jason, Matt, and Brent unpack why traditional truck rolls and “free” phone support are no longer economically viable, why system complexity has outpaced hardware margins, and how recurring revenue programs — modeled on industries like concierge medicine and landscape maintenance — are the only durable answer. Along the way, Matt draws a sharp parallel between the arc of a growing integration business and the moment service stops being a unique selling proposition and starts becoming a liability.
The episode closes on a leadership note: Jason shares OneVision’s data on which integrators actually succeed with recurring revenue programs — and it’s not the ones in the wealthiest zip codes. It’s the ones whose owners believe in the model deeply enough to push through the change management. The conversation is a masterclass in how AI, service design, and operator mindset compound into the kind of momentum that defines the flywheel effect.
Featured Guest
Guest: Jason Griffing
What he does: Director of Product
Company: OneVision Resources
Noteworthy: Jason has spent more than 20 years in the CEDIA channel — 11 years on the front lines as an integrator and 9 years (and counting) at OneVision Resources, where he works closely with CEO Joey Kolchinsky to set the company’s strategic direction.
Where to find her: LinkedIn | OneVision Resources
Key Insights
Context Is the Real AI Unlock — Not Prompting
The biggest leap most operators make with AI isn’t learning better prompts — it’s learning to systematically capture and feed context. Jason describes context as everything from Slack threads and emails to call transcripts, SOPs, and site documentation, and argues that building a personal or company-wide repository of this material is what separates novelty use from transformational use. He walks through his own workflow: at the start of any project, he does a “context dump,” asks the AI to ask him five clarifying questions, then has it produce a markdown document that becomes the foundation for every future conversation on that project.
For integration business owners, this reframes AI from a search-engine replacement into a true thinking partner — but only if you invest the upfront time to give it the information it needs. Without context, AI is generic. With it, AI becomes a compounding asset that gets smarter about your business every week.
The Rock and the Hard Place: Why Service Economics Have Permanently Shifted
ason offers one of the clearest articulations of the structural squeeze facing CI businesses today. The “rock” is rising system complexity — every device is networked, every issue has more potential failure points, and clients expect faster, more responsive support than ever. The “hard place” is hardware margin compression: the categories that historically subsidized free or undercharged service work (TVs, control systems, networks) have seen prices collapse. The result is that the old model — where integrators absorbed service cost in hopes the next project would pay for it — is mathematically broken. For executives running integration companies, this means service can no longer be treated as a cost center or a relationship favor. It must be priced, productized, and operationalized as its own profit engine, or the entire business model erodes.
Recurring Revenue Works Across Markets — What Matters Is Owner Conviction
One of the most counterintuitive insights from OneVision’s data: success with recurring revenue programs has very little correlation with whether an integrator serves coastal luxury clients or mid-market homeowners. What predicts success is whether ownership and leadership genuinely believe in the model and are willing to push through the initial change management friction. Jason has watched mid-market and retail-storefront integrators thrive with premium support plans, while well-positioned luxury firms fail because leadership half-committed. For owners weighing whether their market “can support” recurring revenue, the more honest question is whether they can support the cultural shift internally. This insight reframes service transformation from a market-positioning question into a leadership question.
Episode Highlights
The “Main Character Syndrome” Trap Every Operator Falls Into
Timestamp: ~02:00
he episode opens with an unscripted exercise: Brent asks Jason — a longtime friend of VITAL — to describe what VITAL does in his own words. Jason gets about one-ninth of it right, which becomes a launching point for a sharp observation about how business owners overestimate how well the market understands them. Jason names the phenomenon directly and explains why operators in their own echo chamber consistently miss what’s obvious to outsiders.
“We talk about a concept internally sometimes called main character syndrome… where you’re in your own little bubble as an operator in your business, and you tend to think in these subtle ways that the world sort of revolves around you, and that everybody out there has a great understanding of what it is you do and the problems that you solve.”
How OneVision Uses AI to QA 100% of Support Tickets
Timestamp: ~08:30
ason walks through how each department at OneVision has identified domain-specific AI use cases — and shares a standout example from the support team. Traditionally, contact centers rely on random sampling to monitor quality. OneVision has built an AI-powered quality assurance system that now reviews 100% of inbound support tickets, dramatically expanding visibility into service quality. This is a tangible example of how AI transforms what was previously economically impossible into routine practice.
“Our support team has built out… a new AI-powered quality assurance or QA mechanism that now allows us to perform quality assurance across 100% of the support tickets that come into our contact center, which is amazing.”
The Markdown Trick That Solves AI’s “Bucket Overflow” Problem
Timestamp: ~16:30
Jason gives a clear, jargon-free explanation of how large language model “context windows” work — using the metaphor of a bucket that fills up as a conversation gets longer. He then shares a practical workflow: when a chat gets long and AI starts behaving erratically, ask it to summarize everything into a markdown document, save it, and start a fresh chat pointed at that file. Over time, this builds a library of project-specific context that compounds in value.
“If you’re deep into a chat, sometimes you can see things start to go sideways. That’s because the bucket is overflowing… One of the other tricks with context is, if you get to that point… sometimes a really good thing to do is to stop and say, ‘Okay, take everything that we’ve just talked about and put it into a markdown document for me.'”
The Concierge Medicine Analogy That Reframes Recurring Revenue
Timestamp: ~40:30
Matt shares a personal story about switching to a concierge medicine subscription after a life event — and uses it to crystallize why recurring revenue works for clients as well as integrators. The subscription doesn’t just monetize service differently; it changes the entire experience. Clients get access, familiarity, and speed. Providers get predictable revenue and the operational headroom to deliver on the promise. The parallel to integration service plans is exact.
“Since I’ve been paying every single month, A, I don’t feel this weird feeling about getting a big bill, and B, I can reach them at any hour, any day, text, phone or email, because what they’re exchanging is access.”
