Bo Bennett On Agent Outreach Launch

HOST: Welcome back to Archieboy Holdings News—Bo, it's great to have you back. Last time you dropped a line that I keep coming back to: "It's not very difficult to publish a book—it's difficult to publish a *profitable* book." Everything you've built seems to be chasing that profitability gap from a different angle each time. Today we're looking at the newest one—AgentOutreach.io—which looks like your answer to the part of that gap that lives on the marketing and promotion side. So let's get into it. The site says the AI "figures out who would actually help promote your site," vets that a real contact path exists, and drafts the pitch—but you hit send yourself. Walk me through what that looks like for someone doing it for the first time on a Monday morning. GUEST: Yeah, you bet. It's actually a very easy process. And it's something that most people should be doing when they market their website or their book or whatever. Uh this one isn't specifically for book promotion. It's for any kind of business promotion or website promotion. So, when you're looking for somebody to help you, like, okay, you have a website, now what? You could you could pay for it. You could you could pay Google like $2 a click or whatever to get some traffic, but that gets really expensive really fast. So, one of the best things you could do is find other websites that you can partner with in a way, or you could you could share ideas, or would be happy to promote your service because it would be very valuable to their members. Now, this is a really good idea, but it's a lot of work, and most people just don't bother doing it, especially if you're running a bunch of sites. So what so what our website does, Agent Outreach, is it actually does all that for you through AI. There's two processes. It actually gets everything, it goes out, it does the research, it presents it to you. And then, if you actually have it on autopilot, it'll actually do the outreach for you. So, you don't really have to do anything. So, you talk about a Monday morning, well, you you set the thing up and you forget about it, and then you just wait back and get the responses. HOST: So, it sounds like there's this foundational step, this category planner, that looks at your URL and your description and figures out podcasts or directory submissions, basically, what kind of promoter you need. How accurate is that in practice, and do you get things wrong sometimes? GUEST: Yes, because if you are using AI, AI is uh probability based, you know, in a sense. So, what we do is we do a grading system for everything that AI comes up with. So, it may give you like an A plus um uh prospect, and it may give you a B minus prospect. So, the grading system is based on how good that really is. The human factor that goes into it is you, the customer, that would go through and you would choose to look at each one in more detail and say, "Okay, yeah, let's contact this one. Let's skip this one." Um so, we've we've got like a two paths there. But remember, I mentioned that there's the autopilot, and the autopilot skips that human factor. So, what do we do there? There seems like a conflict. Well, with our initial AI, we train it to be extremely liberal in the sense that it will get as many good contacts that are decent as possible. And may many of them might not be very good. Uh but the second AI, that autopilot one, that's the conservative one where you are allowed to set the grade, and you say, for example, "I only want I only want A minus and A graded uh prospects." So, it's going to be more accurate. So, um so there is the two level of AI, and then there's also you, the person that gets to choose that that buffer. Like if you're okay with sending some emails that not be may not be related to more people, or if you want uh just to go after customers that are very specific and are perfect matches, then you would keep it at like grade A. So, there's a there's a lot of a lot of wiggle room in there. HOST: You mentioned this liberal AI that gathers a broad range, and then the conservative AI for autopilot, but you also allow people to skip leads, and you call it a "skip reason learning loop." Does that mean the AI is actually learning over time based on what users reject, or is it simpler than that? GUEST: No, that that's a it's both actually. Yes, that's exactly what it's that it's doing and it's pretty simple. The uh the AI will collect everything that people skip and it will do like pattern matching and it'll it'll reason and find out, okay, well, why was this skipped? And then in the future, it ignores more of that because if people are constantly skipping a certain type of of uh lead, then it it learns and it will just skip it in the future. HOST: Got it. Now, sticking with that "Monday morning" idea, the final step is this pre-drafted MAILTO pitch that pops open your email client. You specifically call this "the MAILTO advantage." What made you draw the line there rather than just having the tool send in bulk, and do you worry people will see that as a limitation? GUEST: In terms of the limitation, I I I hope not. I don't think so. It is unique that we have we're using their mail server, and we can do all this because of that. If we were to do all the mailing ourselves, we'd have a serious problem with people taking advantage of the the system and doing spamming. And what that means for all the customers, if we were to do that, is their mailing will be on all these blacklists as well. And their uh deliverability of the mailing would only be as good as the worst customer of ours. So, it's it's not a good position to be in, and a lot of companies will do this and not tell the end users about it, and their mail's just not very deliverable. But with this system, we use the person's personal mail, or Gmail, or whatever, and you are basically mailing and establishing your own reputation for mailing as well. So, it's much more reliable, and you could be sure that if, like if somebody complains about a spam, you could address it, and your relationship with Gmail or your mail provider can can stay solid. So, it's just a way for the mail to be more deliverable. HOST: Now, last time you said the whole Archieboy stack exists because it's very difficult to publish a *profitable* book. AgentOutreach seems to live at exactly that profitability gap—the promotion layer. Was this built primarily for authors, or is the real target anyone running a small site who needs outreach? GUEST: The real target is anyone. I actually did build something similar on selfpublishing.pro for authors specifically. It's a very basic version of this and it's just kind of like a one-time shot that uh that gets a whole bunch of connections and maybe you can get like anywhere between 10 and 30 different uh possible connections to make and you just do it on your own. So that that's something that I built for authors. I'm going to test that out. But this like if for any serious authors, this would be really good. But I built this with everybody in mind. The general business. It could be the grocery store down the corner or it could be like a major corporation that uses that has like an international website or it could be an author. It doesn't matter. It'll work for everybody. HOST: The free tier gives five leads a day, permanently. That's a real commitment. But the product's value proposition rests on lead quality—you're promising that each one has a verified contact path and a fit grade. How does the vetting work, and what's your honest answer when someone asks "is this just a scrape with a nice interface?" GUEST: Right. It's uh it's more than a scrape because a scrape can't reason, and that's where AI comes in. AI can reason. Uh with what we're doing is we are, yes, we're looking for email addresses, but we're looking for valid email addresses and we're using the reasoning process of AI to look at the page and see if inquiries are welcome. And that's the big difference. We're not just going after like any email address that's there. It's got to be something that uh like partnerships welcome or or um inquiries welcome, or some kind of invitation that makes it clear to eliminate or at least drastically reduce this idea of spamming somebody else. So you're really more in the line with, hey, I'm looking to create a partnership with this company, and this is the right path to do it, versus I'm looking to spam as many people as I can. So that's where AI AI comes in. And then there's also the form element of it, too, and this is where it really shines, because normally, most websites will not have just an email address, they'll have a form you have to fill out. So what our site does is it gathers all those fields, it'll present all the data based on what the AI knows about the customer's website, and it'll fill out that form for them, and it could submit the form if if the autopilot's going on. And that just makes it a fantastically easy and simple process. HOST: Picking up where we left off—that distinction between a basic scrape and genuinely reasoning with AI, especially with those forms—that's a massive difference. Before we go, for listeners who want to follow up on what we covered, where can they find you and the work you're doing? GUEST: Well, we have two websites that you should take a look at. One, of course, is the site we've been discussing, and that is our agent site. Um, oh boy, what's the name of the site we've been talking about? Agentoutreach.io. Forgive me, I've got so many domain names that it gets confusing. It's agentoutreach.io. And the other one, of course, is our main website, archieboy.com. HOST: Understood. AgentOutreach.io and Archieboy.com. Bo, thanks again for coming on. That insight on deliverability being tied to your own reputation, that was a really key point. And thank you for spending part of your day with us. Let's get you back on soon—there's plenty more we can cover. Until next time—that's a wrap.