Adam Sand (CEO, Roofing Business Partner) demonstrates an RBP-built multi-agent voice AI that automates homeowner intake calls, confirms visual address data, checks weather and drive-time availability, and books appointments directly into HubSpot. He argues this recovers missed leads, frees inside-sales time for higher-value work, and produces measurable ROI using a client example: rescuing 61 stale leads and converting hundreds of digital leads into booked appointments. He presents specific numeric scenarios (leads counts, conversion assumptions, per-appointment pricing) to show a multi-month and annual net-profit uplift, and proposes usage-based pricing ($20 per appointment) tied to results.
Questions this video answers
61 leads
60% (→ 36 retained opportunities)
9 more roofs sold
$162,000 annual net profit created
369 leads
Chapters
Did you know your inside sales team is missing leads and it's costing you real money? I'm Adam Sand, CEO of the roofing business partner in this video. See how RBP's new AI voice call Agent can transform your roofing business by booking appointments, automating intake calls and recovering $162,000 in lost profit, all without hiring a single extra person.
So when I go looking inside of a client account, this is last 30 day leads inside this account. These are more than 30 days old. This is all like March leads. So. So if there's somebody still sitting in marketing qualified, that means that the deal never got to the point of an appointment being set. And if it's lead, it means we never really got a hold of them. There's usually some leads in here that were valuable, we just never got a hold of them. Well, I want to figure out what is the return on investment of deploying voice AI in the business. Why are we not doing that? Like why are we not getting a hold of them? Well, there's two reasons. You have inside salespeople who are not getting back to people, right? They're not getting to the leads fast enough. The challenge is, is that the inside sales reps are not booking the leads as quickly as they could because they're following up with the outside sales reps, they're following up with the customers. They're doing their job file playbooks, they're doing other lead intake during surges of inbound activity, right? Like the phone's ringing, if the phone's ringing off the hook, they're not getting to the inbound leads that are coming through Google. My business, like all those like digital lead creation sources, right? You're hoping that they get the notification. You can create Slack and you can auto text people back, but it's not moving the customer meaningfully towards solving the problem. This is something that every single roofer that we talk to, it's always an issue, it's always a challenge that they're dealing with. And so there's all these voice AI tools that we were going to use. We looked at Charlie, we looked at Synthflow, we looked at Twilio, the go high levels, but all of them wouldn't deliver on the process that we expect our customers to go through. So we had to build one. And so this is what we ended up building. So when that homeowner inquiry comes in into HubSpot, we had to build a multi agent thing is just a way to use AI differently. It's more in line with like, what's happening now in 2025? Which is like. Agent 1 does a visual address confirmation. So it actually looks at the house and says, okay, what is the color siding? What is the color trim? What is the color of the door? This was a whole challenge in and of itself, trying to get it just to get a bearing. This has taken us like, 20 different tries to get right. Because Google Street View doesn't just automatically pull up the house. So getting it to actually, like, travel through Street View and find the house and then take a picture of it and then analyze it, that was a whole problem to solve. Then the next agent has to say, what is the upcoming weather? We're gonna have two days of rain coming up. We don't wanna book three appointments on a day that's raining that we then have to reschedule it, right? So Agent 2 then kind of says, narrow down the days. And then this is the next challenge was using Google Maps API and another agent to say, all right, get the driving distances. Essentially set up Google Navigation to navigate to all these houses from all these houses or to all these houses from this current request and figure out what slots would be available so that it's either four hours away or 30 minutes away and narrowed down by the weather. And then it's got to call the homeowner and it's going to follow the RBP like lead intake script that we have in HubSpot and fill out those properties just like it does in HubSpot. Because we don't want to decrease the impact that we're having with the customer. We want the customer to feel like they're moving meaningfully towards solving the problem. How that essentially run, like, how we expect. Like, how do you calculate ROI on investing in artificial intelligence business? That's like the next piece. Like, if we're consulting you on how to use AI in your business, how do you make a determination as to whether or not AI works? And so this is something that we work through. So that client, I pulled two lead sources. It was Angie and Google, my business, right? Where there was 61 leads that month that never got to sales, qualified lead, or other words, never got to the point of being an appointment booked. So lead came in, we didn't see it right. We didn't get to it in time. If we were to book 60% of those, that would be 36 retained opportunities. If you get 60% of those leads to become retained opportunities at your current closing rate, that's nine more roofs sold or 162,000amonth. Now we have to think about that in terms of net profit. So at 10% net profit, that's $16,000 in a month. So then I'm thinking, well, how does that work out per year? Now this client doesn't work 12 months a year, they work 10 months a year. So then I had to say, okay, well that net profit times 10 months a year is $162,000 in annual net profit created. But that's not the only thing. There's going to be a reallocation of effort. Like when you're deploying AI, you're not doing it so your employees can like put autopilot on and kick back and play foosball and goof off. No, you're doing it so that you can remain competitive. If you don't do it, someone else will. Like the time to trick everybody into thinking that you're not using AI when you are and then put your feet up and let AI do the work for you. And nobody's gonna know. Like that was a year ago, everybody knows it exists now. Like there's no surprise. This is about retaining a competitive edge. So then you have to reallocate effort to revenue generating activities. So they had 369inbound digitally created leads that month. That means that the lead came into the CRM that this agent can call and book those appointments. If you take 369, 10 minute phone calls and make it so those do not have to happen. In addition to these 61 phone calls here that never happened. Well, those 369 phone calls plus that, like if you don't have to do those lead intake phone calls and you're still giving a great experience, well then you're going to increase your close rate from 25 to 30. Right. If you just go from a 25 to 30% close rate, you're going to see an increase in number of deals sold because you're reallocating your time from hi, what's your name, what's your phone number? Da da, da. Like the receptionist eventually gets lazy. If you take those 369 calls and still give the customer a great experience, train it on the voice of the inside salesperson so it even sounds the same. But then just reallocate their time to post inspection, debrief, call, follow up call, they're going to sell more. That's 61 hours a month spent on lead intake calls. So if you put 61 hours back into the inside sales team for follow up, you're going to make more money plain and Simple. So you're going to have more closed roofs. And then what I wanted to do was say instead of charging like all these SaaS companies where it's just another thousand dollar a month cost, I want to charge for a result, charge for a result. So cost of operations. So then $20 per appointment booked we get of those 369 leads, we booked 295 created appointments. So then we would charge for just the created appointments. So that way the roofing customer, if they're slow in the winter, they're not paying for this thing, they're only paying for usage. So if you're going to do this, you're only paying per appointment booked and then you're reallocating that. So this would be like 77,000 per year, right? That they would pay for this. And this is how that return on investment would go. In the first quarter you'd pay the 2500. But you're getting these appointments and these sales booked and then this is continue to go. So this should get a 2 1/2x return after year one. But that's not considering that. If I add this back into the calculation saying, hey, what about the reallocation of effort? Well, if I add that, then it's a 6 and a half X return. But this is all based on real customer data. And so this is how we have to start thinking about AI. How do you compete? How does a company who doesn't do this compete in this realm?
