Frideric Prandecki

One of the largest HVAC company reps in my market just texted me this: “The biggest financial crisis since the Great Depression is underway.” Not a hedge fund manager. Not a pundit. An HVAC rep. Meanwhile others think we’re at the start of a new bull run. This feels like 2007 — when nobody agreed on anything.

One of the largest HVAC company reps in my market just texted me this: “The biggest financial crisis since the Great Depression is underway.” Not a hedge fund manager. Not a pundit. An HVAC rep. Meanwhile others think we’re at the start of a new bull run. This feels like 2007 — when nobody agreed on anything.

Week 1 down! On Monday, we rolled out AI to the front lines of our contact center. I was nervous. But we measured the results, and I'd give us a B/B+. This change has come with one non-negotiable: we better measure what matters. As we reimagined our call center, we introduced two new KPIs--both of which are crucial to ensuring that responsiveness and quality don’t suffer in an AI-first model. 1. Speed to Answer (Goal: <10 Seconds) Here’s the scenario: A customer calls after hours. Our AI agent answers immediately, understands the issue, and escalates the call to a human agent based on predefined criteria (e.g., human request, frustration, or nuanced requests). Speed to Answer measures how long it takes for our human team to pick up that escalated call. Our target? Less than 10 seconds. If a customer needs a human, they shouldn’t wait. This metric is simple, precise, and an instant proxy for whether our AI + human handoff is working the way it should. 2. Speed to Resolution (for AI-Flagged Follow-Ups) Not every call ends neatly. Some are disconnected. Others involve objections around pricing, scheduling, or service availability. That’s where Speed to Resolution comes in. This KPI measures how quickly our human agents follow up on leads that were flagged by AI as needing further attention. We’re still refining the targets here--but we know not all follow-ups are created equal. A customer requesting an invoice doesn’t need the same urgency as one who’s sitting in a 40-degree house waiting for heat. So we’re building tiered resolution goals based on the priority level of each flag. We’re monitoring these KPIs in near real-time to ensure that the move to AI doesn’t come at the expense of quality or care. We’re bullish on what AI can do--but only if it’s paired with human empathy and operational discipline. The results so far? Near-instant response times during peak call volume Improved consistency in data collection Fewer dropped leads Better prioritization for dispatch and scheduling ... and the contact center is just the beginning. Our belief in AI isn’t abstract. It’s operational, measured, and it’s evolving. The call center was the first step in a broader AI transformation that will touch scheduling, dispatching, capacity management, and customer follow-ups in the months ahead. We’re not replacing humans. We’re amplifying them so they can focus their energy where it matters most. If you’re running a home service business and looking for a way to scale without sacrificing service, the future isn’t in more headcount. It’s in a smarter system. We’ve just started building ours...

US consumer delinquencies are surging: 3.0% of auto loans transitioned into 90+ days delinquency in Q3 2025, the highest in 15 years. At the same time, 7.1% of credit card debt became seriously delinquent, near the highest in 14 years. Student loan serious delinquencies spiked +13.5 percentage points YoY, to 14.3%, an all-time high. This followed the expiration of the student loan relief period, as missed payments began reappearing on credit reports. Mortgage delinquencies also increased, with 1.3% transitioning into serious delinquency last quarter, the highest in 8 years. Consumers are drowning in debt.