The B2B Podcast Index
AI for Business Podcast

E52: How One AI Campaign Generated $200,000 From Dead Leads featuring Bryce Decora

AI for Business Podcast · 2026-06-19 · 23 min

Substance score

39 / 100

Five dimensions, 20 points each

Insight Density7 / 20
Originality7 / 20
Guest Caliber12 / 20
Specificity & Evidence11 / 20
Conversational Craft2 / 20

What our scoring noted

Our reviewer’s read on each dimension, with quotes from the episode.

Insight Density

7 / 20

The episode contains a handful of genuinely useful ideas - the danger of over-automating big-ticket sales, the human-trust gap in AI, and the flow-diagram vs. text-prompt approach for scale - but the majority of the runtime is consumed by origin story, crowd engagement theatre, and a product features pitch that offers little transferable learning.

it's really easy with a typed out type bot to get a really cool demo that can impress someone, but it doesn't work at scale
Trust doesn't happen with AI. Trust happens with people

Originality

7 / 20

The personal cautionary tale about destroying his own real estate business by over-automating is a refreshing counter-narrative for an AI conference, and the 'recession of trust' framing is mildly interesting, but the attract/engage/delight flywheel is a well-worn framework and most of the AI-automation advice circulates widely.

we're in this day and age where there's this recession of trust
the easiest thing to do is be a person

Guest Caliber

12 / 20

Bryce Decora is a genuine practitioner - he built prototype AI at Boeing in 2018, applied it to his own real estate operation, and founded a product with verifiable scale metrics - but the conference-pitch format constrains depth and the talk functions primarily as a sales presentation rather than knowledge transfer.

this was 2018 when I started working with AI ChatGPT came out in 2022. So we were training our own models to do image recognition and search airplane drawings
It's processed 100 million messages across 35 million contacts. And it's booked. Now that's wrong. It just crossed 600,000 booked appointments

Specificity & Evidence

11 / 20

There are multiple concrete numbers - 600K appointments, 100M messages, 35M contacts, 60 years / $5M saved, and the Joey Brown roofing case study with 773 leads yielding ~$200K - which is above average for a conference talk, but all figures are self-reported marketing claims with no external verification or methodology disclosed.

Joey Brown...773 leads that this roofer said were dead...he generated almost $200,000 in revenue for this roofing company by doing one reactivation campaign
we estimate it's saved businesses about 60 years of time...which is about 5 million and saved money

Conversational Craft

2 / 20

There is effectively no host craft to evaluate: this is an uninterrupted conference sales presentation, the host contributes nothing until a three-second outro, and there are zero follow-up questions, challenges, or probes of any claim made.

That's a wrap for this episode of the AI for business podcast. If you enjoyed it, don't forget to hit subscribe and pass it along to a friend or colleague.

Conversation analysis

Computed from the transcript - who did the talking, and the verbal tics along the way.

Filler words

like34so33right7you know2literally2I mean1sort of1kind of1actually1anyway1

Episode notes

Bryce Decora is the founder of CloseBot, an AI-powered lead qualification and appointment booking platform that has processed over 100 million messages across 35 million contacts and booked more than 600,000 appointments, saving businesses an estimated 60 years of collective time. Before building one of the most-reviewed AI tools in the space (20 G2 awards in a single season), Bryce was a mechanical engineer and software developer at Boeing who accidentally automated himself out of a real estate business and learned the hard way exactly where AI belongs in a sales cycle. In this session from the AI for Business conference, Bryce shares the framework he built from that failure: the three-phase business flywheel that separates businesses printing money from businesses burning it. He breaks down where AI should handle the heavy lifting, where human trust is non-negotiable, and how one ClosedBot power user turned 773 dead leads into nearly $200,000 in revenue with a single reactivation campaign.

Full transcript

23 min

Transcribed and scored by The B2B Podcast Index.

What is up? This is such a cool crowd. I did the panel earlier this morning and everyone was just so engaged. I'm used to speaking in front of people that I don't know what the deal is, but they're not like a family. You guys are so cool. So I want to do a little experiment and bear with me. I want everyone to stand up. Stand up. You can do it. Everyone stand up. I'll wait. Stand up. If you love eye for business, how about that? That'll get them. Yeah. Okay. Okay. Yes, that gets them. Okay. Now I want you to sit down. If. Sit down. If you are just starting your business, like. And just starting something. Now sit down. If you've been in business one year. Sit down if you're an agency. I'm just trying to get to know my audience. Sit down. If you have not used your closed board account. Yeah, I know. I joked about this in the panel. I looked at the numbers before coming and there were like 3000 people with accounts, and we have like 76 people using it, which is wild and also shows how much opportunity there is in doing this. So I don't know if you guys know my background. I grew up in the Midwest. I went to engineering school and I moved out to work at Boeing doing mechanical engineering software development, and I and my parents were so proud. Can you imagine growing up in Nebraska and your little boy goes to work at Boeing? But they didn't know how unhappy I was working at big corporate America. And many of you have probably felt the same way doing something in the past. And maybe you're stuck doing that right now and you're here taking vacation time, sick leave to try to break out. And I want to start by saying, you guys can do it. I did it, and you can do. I finally did it by watching my wife and what she did. She's out here somewhere wearing the clothes bought shirt. She was a realtor. And even though I was a software developer working on prototype AI projects for Boeing, she started consistently making more money than I was selling real estate. And I was like, hey, that's cool. B that's not fair. So I went on sick leave like some of you guys may be doing right now, and I decided to try out real estate investing, and that was my first glimpse into sales. And my God. Was it hard being a tech person, liking to fiddle with all the latest tools, write code, sit at my desk and not talk to anyone? And now is my job to be on phone calls and texting eight hours a day, every day with distressed homeowners. And I was like, I don't have eyes and I shouldn't have to do this. No one should have to do this first of all. But I think I can I think I can program my way out of this problem. And I started to talk to my other friends about it in real estate, and they were like, you're crazy. Can't be done. And when someone tells me it can't be done, I'm like, sure, bet. Okay, let's go. So Laura can attest to this. I was not good at making sales cause I was spending all my time working on automations and code and AI. But I knew on the other side of this was the dream. The green pasture of never having to talk to anyone again and just having money come in. I knew I could do it. So I built and at the start I had like a hybrid of AI and Vas. As time went on, less and less Vas. People still told me I was crazy and I was like, I'm close. I feel it. And then it happened. I made a $75,000 commission check and I was like, all this is going into AI automations, and I'm going to turn the knob all the way up. And guess what happened? I proved all of them wrong. Thank you. And I drove us out of business. Yes. So I was like, great. What do I do now? I'm in the similar sort of boat. I was at Boeing, except now I had nothing to do. No goal to go to. And then some people from real estate came to me and they were like, first of all, I told you so. Second of all, I do think you have something really powerful, and I would like to pay you for that thing. They were like, I'm not going to do what you did. We're not going to go all in. We're going to leave in the human component where it matters. But we can use that to qualify our leads, like I was doing at the beginning, and then bring in humans later. And so now we have closed bot what it is today. It's processed 100 million messages across 35 million contacts. And it's booked. Now that's wrong. It just crossed 600,000 booked appointments. Thank you. It's really cool. This stuff is really working. So you may still be thinking, okay, why use AI? A lot of you guys out there are probably doing lead qualification yourself. You have virtual assistants. Maybe you have some automations, questionnaires, and it's a valid question. Even now, why use AI as an appointment setter? And let's go back to that. 600,000 appointments. When you look at that and you take into account how many messages have been processed and how many leads have been disqualified. How many questions have been answered? How many appointments have been booked but also rescheduled and canceled through AI? We estimate it's saved businesses about 60 years of time, and that's if you look 24 seven, 365 days a year, 60 years, which is about 5 million and saved money. But the most important thing is elevation of the workforce. Now, this may be you. This may be your virtual assistants, but scheduling and qualifying leads that may or may not be worth your time is literally the worst job I know, because I did it eight hours a day for a long time and it drove me crazy. And you guys probably feel the same way. And I did two. And so that's the part that we're trying to get to. When you look at single messages coming in and a lot of you business owners may be dealing with this right now, it's pretty manageable. You get maybe you're doing a database reactivation campaign, reaching out to clients and a response comes in easy enough. In this case, tagging is not interested in put your phone down. Go back to family watching, watching the big game. You know, whatever. Okay, another message comes in a little bit more complicated, but still, we're just dealing with one at a time. This is fine here. Let's maybe update our CRM a little bit, update what insurance they're using. Try to book them for an appointment. But where it gets hectic is where we start to have maybe two come in at the same time. You're working out of one inbox. Okay, now you've marked that is red on accident, but you don't want to respond yet. You go into a different one or you you forgot to respond to that other one. Okay, now more are coming in. Okay, now it's 6:00. I'm with family. I'm trying to eat more messages are coming in. You're stressed out because you're with family. But you know, if you don't respond to these leads in 10s, that's what they tell you. Respond in seconds or it's pretty much gone that you won't make money. You need to feed your family, and it just gets worse and worse and worse. And that's where close back comes in. As the problem we're trying to solve for you, it's not to completely remove the human component. I called it close by when I intended to have it go for start to close. But that's not what is designed to do. It's meant to close the appointment for you guys. But how do we do that? Well, I saw early on at my time at Boeing, this was 2018 when I started working with AI ChatGPT came out in 2022. So we were training our own models to do image recognition and search airplane drawings. And it I was not good at all. So you really had to guide it. And this is still kind of true today. I really likes to go off this way. If a contact starts talking about something, it loves to just follow their train of thought and this keeps it on track. So in this example, this is a basic just question and answer agent within close bot. It's telling it. When a conversation starts I want you to sit there and answer questions that maybe we want to answer questions. But if we detect they're unhappy or not interested in our product, we want to tag them in our CRM accordingly and then stop responding. Similarly, if we're sitting there answering questions and they do show interests, we might want to collect, in this example, the type of insurance they have, updating the custom field in high level or HubSpot. Raise your hand if you're using high level. By the way. Yeah okay. Perfect. So we might want to update those custom fields right. Maybe update the timeline on when they want this project done and ultimately book an inspection appointment to your calendar. This keeps the AI controlled at volume scale. It's really easy with a typed out type bot to get a really cool demo that can impress someone, but it doesn't work at scale. Just like with your virtual assistants or yourself, if you're mapping out your thoughts, what's easier to follow a diagram or big paragraph of text? What do you think works better? Diagram? Yes, and that's what we found early on, and that's what we've leaned into more and more. So remember at the beginning I talked about how I killed my business in real estate? Sorry, Laura. My wife. I'm sorry. Again. We had some hard times going through that, but because of that, I was able to learn where AI fits into a business. And that's what I want to share with you today. Who's heard of a flywheel? Raise your hand. Yes. Business. Flywheel. That's what we all want to get to. That's the point where your business is literally just printing money. And to do that, we have three phases of the sales cycle. Attract, engage and delight. And if you do these rates, your business feels super easy and low stress. If you do it wrong, it feels like you're continuing to throw money and energy into your funnel. And as soon as you stop, your business shuts off. That's what we don't want. Close by fits in the attract phase. So those of you who have plenty of leads coming in and you need a way to qualify them, or maybe people coming in after hours and you need to qualify them, that's the attract phase. And that's where close bot fits. We still have two other sections. That first one is the attract phase. That's where closed bot is designed to go. Then we need to think about your customers and where their first win happens. For most people that win involves trust. Trusts doesn't happen with AI. Trust happens with people, and that's going to become so much more relevant as time continues to pass and we get more and more people like me several years ago, trying to automate the whole thing. And I don't want you guys to fall into that boat. Now, let me preface this by saying, sometimes the the win that your customers are looking for can be fully automated, but in a lot of outbound scenarios or big ticket items like coaching or a home sale, they need a person there to make them feel heard and understood, and close by is designed to expedite that customer's path to that person that they can trust and get their first win with. So engage human trust. And then at the delight phase, that's where you have to fulfill. If you're doing real estate, that's the whole transaction coordination process, which can be a combination of human trust in automation. And if you do that part right, that's where they tell their friends about you. That's where they proclaim how great you are on the internet. And then more leads just keep coming in. And it's the weirdest thing. You start getting people that are just like, hey, I want to, I want to buy, and then they buy and then they tell more people about you and it just happens. Raise your hand if that seems cool or don't you want that for your business? Now what you don't want. Is this. And this is what I accidentally did. This is what you don't want to do. Lead comes in. Use closed bot. They want a person. You continue to use closed bot. They scream for a person. You continue to his close bot or some other tool, some other AI, some other automation flow. And they tell all their friends, hey, yeah, don't use that company because they don't care about you. And we're in this day and age where there's this recession of trust. So it's so exciting because us as people, the easiest thing to do is be a person. It's so easy. The hard thing, especially for all of you. I've talked to some of you, the hard thing is the automation side. And so you may be feeling overwhelmed at a conference like this where we're talking about all these cool tools and lovable and close by and AI employee. But the coolest thing you can do for your customers right now is be a really good person. Maybe pick 1 or 2 tools that resonate with you and can help you expedite that customer to trust, and then just be a really good person. It's so easy. If you're overwhelmed, I want to say you are in the right place. I'm not saying feeling overwhelmed is the wrong thing to feel. That means growth. We're all growing here, okay? And the cool thing about AI for business is they're here bringing not only bringing you the best tools and sharing all this knowledge. I mean, Brian doesn't sleep, so he's learning all these tools for you, and then he shares what's good and not good. So lean into that. But they also have this whole fulfillment team. So those of you who have a closed board, you may not even know you have a closed bot account. You do in V1. We're in V2 now, and we're shifting into this phase where you have two options. The umbrella is you get to use the best text based AI out there, which is closed bot. You can either get your own account and we have a discount code. I'll show. But you can also just contact AI for business and they can build for you. They have a bunch of pre-built agents that they can plug your high level account into. You just let them know, hey, I want, a support agent or I want to lead qualification agent for this scenario. They have all of it built out already. For most of you, you have enough to worry about, and you just need to contact Brian and Francis and say, hey, yeah, hook me up with that. They have it all built. That starts November 10th. Those of you with calendars, that's Monday. That's coming soon. This moves fast. And just know that either way, no matter which one you choose, you're working with the best. If you look us up online, you'll see on G2 we have I think it's like 90 G2 reviews saying it's not just another chat bot tool. This actually works. G2 gave us 20 awards during their fall awards session. We have thought of things to make your life easier like the smart FAQ. I was talking about it a little bit on the panel. I was on our goal. Our vision is for you to have the AI qualifying leads, and for you to never have to review conversations. This was a step in that direction. The AI will notify you when it doesn't know the answer to a question. You don't have to go in and review the conversations, you just have to click. Fill in the answer and click save. And then it and it knows it for the future. That's how it's getting smarter over time. You don't have to go and review stuff. We're highly reliable. Since we have launched and stabilized V2, we haven't had downtime ourselves, and that's because we use all of the main AI providers you get to pick. Do you like how grok sounds? Do you like OpenAI, AI, anthropic, Gemini, Deep Seek? You get a pick, but most importantly, you can set which one you like the second best and third best, fourth. They're not all going to go down. Probably. So that way you're still up and running, and even if it's late at night and you're with family hanging out and OpenAI goes down, you don't have to say sorry, family. OpenAI went down again. I knew this was going to happen. They just released a new model. I'm going to be out of commission for the next 24 hours, trying to catch up with all these leads. That's not something you have to worry about with closed bot. And even just simple things like being able to email most of the AI chat systems out there can't email, and even taking it a step further, if it were to chat with someone and they said, hey, can you just email me your prices, it can fire off that email and continue chatting over text message or live chat on your website. We have the flow builder that I was telling you about. Like I mentioned, I've been really doing the flow built AI stuff since my time at Boeing in 2018, so we have it nailed. And scenario detection. You guys know that conversations aren't linear. I may be chatting with you, and I have a good plan of where I want to go with qualifying you as a lead, but then you ask something weird like, can you send me the zoom link before, or can you send me the recording link for this? By the way, and you might want it to be able to jump down and do something a different job if that's asked. Or maybe you're qualifying leads, but they say, hey, enough questions. I just want to book on your calendar. If you want to allow that, you can have that scenario listening for that and then just jump straight up to booking. So these are really powerful. For it to be able to handle like a human would. And really a lot more. I'm not up here to spew features. Who knows this guy? I'm curious. Joey Brown. It was like one hand, okay, Joey Brown's in your guys's community. He's a close bot power user. He did a reactivation campaign for a roofing customer and 773 leads that this roofer said were dead. He's like, I don't care if they're all pissed off by the end of this. They weren't going to generate business anyway. And he generated almost $200,000 in revenue for this roofing company by doing one reactivation campaign. So if you guys have contacts in the database, you're not doing anything with or know people who do, and you can do a database reactivation campaign for them that is low hanging fruit, friends, and it works. That's a wrap for this episode of the AI for business podcast. If you enjoyed it, don't forget to hit subscribe and pass it along to a friend or colleague. We appreciate you being here and we'll catch you in the next one.

More from AI for Business Podcast

All episodes →
Explore the best B2B AI & Data podcasts →
Listen to this episodeAll AI for Business Podcast episodes →