Episode Details
Podcast: The Jeff Bullas Show
Episode: The Surprising Comeback of Direct Mail: How AI Makes It Better (Episode 247)
Published: January 16, 2025
Guest: Wilson Zehr
Guest’s businesses: Cendix and Zairmail
Guest’s title: Founder of Cendix (Zairmail); Technology Entrepreneur and Academic
Host: Jeff Bullas
Duration: Approximately 44 minutes
Apple Podcasts: https://podcasts.apple.com/us/podcast/the-surprising-comeback-of-direct-mail-how-ai-makes/id1502649184?i=1000684307874
In a world dominated by digital marketing, postal direct mail might seem like a relic of the past. But according to Wilson Zehr, Founder of Cendix (Zairmail) and experienced tech entrepreneur, physical mail is making a strong comeback—and it’s proving that tangible connections still matter.
Wilson has launched six tech startups and raised $60 million in venture capital, giving him a unique perspective on customer engagement and the role of innovation in marketing. Through his work with Zairmail, he’s showing that the right combination of digital tools and physical mail can elevate marketing campaigns to a level that purely digital channels often cannot match.
Why Direct Mail Still Works
Direct mail cuts through the noise. In an era where inboxes are flooded with emails, notifications, and digital ads, receiving a well-crafted physical mail piece grabs attention in a way that pixels can’t. The tactile experience of holding a letter or postcard creates an emotional connection that fosters trust and engagement.
Combining Digital Automation with Physical Mail
One of the keys to modern direct mail success is integration with digital automation. By leveraging technology, businesses can streamline the process of sending targeted physical mail, making campaigns faster, easier, and more cost-effective. This approach allows marketers to focus on strategy and creativity rather than logistics, while still reaching prospects with a personal touch.
The Role of AI in Direct Mail Marketing
Artificial Intelligence is transforming how we plan and execute campaigns. From improving targeting and personalization to predicting which messages will resonate most, AI helps marketers maximize ROI and minimize wasted effort. Yet, as Wilson notes, AI can’t fully replicate human creativity or the emotional impact that storytelling brings. A personalized, thoughtful message crafted by a human still has the power to delight, surprise, and engage in ways algorithms alone cannot.
Lessons from Six Startups
Wilson’s experience launching multiple startups has also shaped his approach to marketing. He emphasizes that past actions and insights can inform better decisions today—whether it’s choosing the right audience, timing campaigns, or innovating in your messaging. Direct mail, paired with thoughtful strategy and technology, can be a vital tool in both startup growth and established business success.
Turning Mail into Marketing Success
Postal direct mail isn’t just surviving—it’s thriving in today’s marketing landscape. By blending digital automation, AI, and human creativity, businesses can cut through digital clutter, build meaningful connections, and drive results.
Ready to see how Zairmail can transform your marketing campaigns? Start sending smarter direct mail today and bring your message into the hands of the people who matter most.
Apple Podcasts:
https://podcasts.apple.com/us/podcast/the-surprising-comeback-of-direct-mail-how-ai-makes/id1502649184?i=1000684307874
The Jeff Bullas Show
https://thejeffbullasshow.libsyn.com/the-surprising-comeback-of-direct-mail-how-ai-makes-it-better
👉 Book on Calendly
Wilson Zehr is a Direct Mail Industry Expert, Direct Marketing Specialist, and CEO of Cendix (www.cendix.com), the leading provider of web-to-print solutions. He is also the founder of Zairmail (www.zairmail.com), responsible for many industry firsts in direct mail.
Free Direct Mail Sample Kit
See direct mail before you send it
Get samples of direct mail postcards and letters, plus service details to help you launch your first campaign. Your kit ships free by USPS.
Request Your Free Sample Kit →Click to View Full Episode Transcript
Podcast: The Jeff Bullas Show
Episode: The Surprising Comeback of Direct Mail & How AI Makes It Better
Host: Jeff Bullas
Guest: Wilson Zehr (Seasoned Entrepreneur, Academic & Founder of Zairmail)
[00:00:05] Jeff: Hi everyone and welcome to the Jeff Bullas show. Today I have with me Wilson Zehr. Now, Wilson is an experienced technology entrepreneur and academic. In fact, he’s got three PhDs. As a founder of Cendix or Zairmail—a company streamlining the process of sending physical mail, and basically, we can call that snail mail today as a bit of a… but basically sending mail for business development—Wilson brings a unique perspective to direct mail marketing in today’s digital age. And I’m going to be curious about what that is for a bunch of things it’s about. While many might think it’s outdated, it definitely still does work. Throughout his career, he’s had the opportunity to launch a total of six tech startups, raise $60 million in venture capital, while maintaining a strong belief in the power of direct mail for customer engagement. And along the way, he’s managed to do a few degrees and three PhDs. So Wilson, welcome to the show.
[00:01:02] Wilson: Thank you. Thank you for having me, Jeff. Um, appreciate the opportunity, for sure.
[00:01:02] Jeff: Okay, so you’ve done more degrees than most people do in six lifetimes, which is just amazing. So obviously you must be curious and also just love learning. So um, but you did a… and now you, we’re, you’re doing this direct mail—in other words, things that arrive in a letterbox, right? Yeah. So I’m going to go straight to that, actually. What made you start Zairmail? Um, because I’m… this is the most curious thing. In a digital age, you’re doing physical—not email, physical mail directed to people’s letterboxes. What inspired that?
[00:01:45] Wilson: Yeah, great question. Um, I’m, I’m a little, uh, perplexed by that myself some days. Anyway, that when, um, we worked on a couple of startups… um, well, the f- first, um, the first startup we did on a, I did on my own was something called Efusion. It was jointly funded by Microsoft, Intel, France Telecom, and Telecom Italia. And what we were doing was vot-, looking at Voiceover IP applications. Yes. You know, some of the technology we worked on is still in use today in things like Skype. Wow. Um, so that, but it was that, at that point in time, you know, when the internet was first, you know, getting traction and exploding. And the internet, by the way, it’s been around for a long time, it’s just the browser and stuff was something new allowed us to connect regular users to the internet. It democratized the internet, didn’t it?
[00:02:47] Jeff: Yes, it did. Yeah.
[00:02:47] Wilson: And it really, I mean, people use the internet in different ways, too. In one way, it’s a big communications pipe. In another way, it’s a data store. In another way, it’s a way to, you know, access information or, or to, to, not access information, but to complete transactions, right? So, um, it does lots of different things depending on how you, how you link it together. But you know, the, the browser and HTML really made that all possible. I mean, I was using the internet for, in, in the late ’80s to send email from a software company I worked at called Verdict. I did the news feed in the Northwest, in the Pacific Northwest, in that period of time. So it was available in the university community, it just wasn’t available to most consumers.
[00:03:33] Jeff: Yes. Yeah, and that, that was intriguing. So for me, uh, looking at the internet was used what, from about the 1970s I think right through to the 1990s exclusively pretty well for the academic community. In other words, it was a academic product, actually.
[00:03:49] Wilson: Yeah, actually, the internet was a DARPA project. It started as a military project and the whole idea was to create a, a network that had no single point of failure. Because when, you know, when you’re attacked by an enemy, you know, traditional theory is that you want to take out, you know, you want to take out communication, you want to take out supplies, and you want to take out power, you know?
[00:04:15] Jeff: Yeah, yeah.
[00:04:15] Wilson: And, and coming up with a, a, a network that had no single point of failure seemed like a very good way to make sure that the country itself was more resilient. So they started out building a network among major universities here in the United States, both East Coast and West Coast. And they had to create, um, some different devices to do that. The original computers weren’t capable of doing that, weren’t even, weren’t capable of talking to each other at all. It’s a company called BBN made those first switches. It’s actually, it’s a really interesting story. There’s a great, um, um, there’s a great video on it on the History Channel where they talk about, you know, they talk about the, um, all the enabling technology. So the, the switches, BBN switches, the creation of TCP/IP, um, ethernet—they had to invent ethernet to transmit signals—and then of course all the servers had to, you know, had to be become a little more intelligent, a little more powerful. And then the end points, right? There’s a lot of, um, changes required on the PC side, too, or the personal side.
[00:05:21] Jeff: Yeah.
[00:05:21] Wilson: Yeah, I’ve got a really great graphic where I, I break it down to the level. So we look over time and how these, um, how they feed into each other. The cool thing about, um, um, diffusion of technology innovation is that it tends to layer on itself over time, you know? One layer after another layer after another layer, and the, the layer before enabling the next. Um, so it’s really interesting how these new, each new technology inspires a whole, you know, a whole wave of, of even newer technology that people build on.
[00:06:00] Jeff: Okay, so let’s get back to the original question. Why are we using old analog technology called direct mail? And I’m, that’s, that’s what I’m really curious about.
[00:06:09] Wilson: All right, so, so I, so to go back to, to where we started, the, there was this, um, there was this recognition at that point for me that um, there was something called convergence, right? So we had all of this legacy, all of these legacy systems that were in existence that were doing jobs, but we now we had the internet and people were counting on the internet to do certain kinds of work. And there was a huge opportunity to, to, to bring convergence between those systems—to somehow connect legacy systems with the internet, you know, internet applications and internet services that people were using. So in Efusion, we did Voiceover IP and we basically, we connected the internet to the public telephone network—the old public phone networks. So we’re dealing with old switches, old analog switches and stuff, and trying to figure out how to turn that into digital voice.
[00:07:05] Jeff: Yes, it’s kind of interesting, um…
[00:07:05] Wilson: And so Zairmail is really, um, focused on convergence between electronic media—and everything starts out electronically today, if you’re doing a letter or a bill or anything, I mean, it doesn’t start out as a piece of paper anymore, it starts out as a digital file.
[00:07:21] Jeff: Yes.
[00:07:21] Wilson: And, and if you have to mail that to somebody, how does that work? And so we’ve created this process that allows you to submit a document, submit a list, and we would, um, we route it electronically to a place near where it needs to go and then produce it and mail it. Um, so it’s, it’s, it’s faster, it’s easier, it’s better for the planet, um, and it just makes sense within the context we work in today. But it’s not the only um, opportunity like that. I mean, we, our first, um, Zairmail started out in the Draper Fisher portfolio. Um, DFJ is a tier-one California VC, and they also were the guys who did, um, you know, they funded Skype, they did um, Hotmail—which was free email, remember that platform?—they did Efax, which was, you know, convergence free faxes, they did Homestead as free websites. So there’s a whole series of services like that in the portfolio, and we just fit, fit pretty well with that offer, you’re right. Um, but Zairmail, the, the concept of Zairmail’s, um, been around for a while. I mean, we started… the first company was first started long ago. Um, it’s been sold a couple of times, we got it back, um, and, and that’s a longer story that I’d be happy to share sometime. Um, but you know, we brought it up on our new technology, started up, and orders kept coming in. The demand is there for the product that we offer. And the interesting thing about what we do, or the way we approach what we do, is most people in our space are commercial printers and so they’re very focused on the production side of things, which is good. I mean, we have partners like that we use in our network, too. But we, we are a software and technology company, and we always have been. So that’s the way we look at the problem, just as we did with Efusion looking at Voice over IP, right? So we’re looking at: how do we, how do we use technology to shape the kind of applications we can deliver? So we have an application called Flights, where we’ll take a big, you know, a million records, and we have documents and schedules, and we’ll automatically slice it up, send it out, you know, track, track, um, you know, track the progress of the mail and, and alert people when it’s been delivered. And that happens lights out, 24/7, you know? we can just start that up and it goes. We, do, you know, we can take a single job, break it up, and spread it across an entire network across the United States. We can do… we’ve got a new one, a new application that’s really kind of cool. It’s, um, it’s what we call Speedy, and anybody wants to can download it. Speedy is free to download. It’s at zairmail.com. And what it is, is a desktop application. You download it, it puts an icon on your desktop, you create a letter—even if it’s just one letter—you drop it on the icon, you, it says, “Where do you want to send it?” You put in an address or select an address from your, you know, from your um, address book, and out it goes—it turns into a letter and gets delivered in the mail, right?
[00:10:35] Jeff: But okay, so tell us about okay, how you actually execute a digital to… do you have companies, a company that actually takes digital signal and then actually prints a letter and then puts it in a, in a…
[00:10:50] Wilson: We do so, we…
[00:10:50] Jeff: …and then puts a stamp on it? So I’m trying to understand. We got the concept. I’m just wondering, take me through what actually happens physically, because we’re talking about physical delivery, and why would you do it and what are the results you’re getting?
[00:11:12] Wilson: Yeah, so we have a, we have a network of production partners today, and these are commercial printers that also do mail. We have partners today in, you know, in Oregon, um, two locations in California, in Dallas, um, Maryland—you know, Baltimore, Maryland—and in Iowa, yeah, um, and we’re always looking at tuning the network, changing the network. Um, but the basic idea is that, uh, customer will come to the site, they upload a document, they upload their list, um, they see a preview in real-time of what’s going to be produced, and they like what they see, they push the button, and they’re, they’re in the mail tomorrow, the next day. So really, literally takes 60 seconds to, to um, launch a campaign. We’re looking across the network, we see: what is it being produced? You know, what, what facilities have the ability to produce that file or that job? So it could be a postcard, or a letter, or a self-mailer, snap pack, or, you know, any number of things—could be black and white or color. So we’re looking across the network and see who, who has the ability to print something like that, um, and then where’s the most effective place to print it? So for, so for example, you would say, well, maybe this letter is going to, you know, so maybe it makes more sense to print it in Baltimore. Well, but if it’s 3:00 in the afternoon here on the West Coast, that doesn’t make more sense because, because they’re not going to print it today. But if I started in Sacramento, I could print it there, and I’d already be, you know, in the mail before the other guys even started on it. So we have, we, you know, use some heuristics like that to figure out the best space to produce and mail. We also provide tracking on every piece of mail, so we can tell how it flows through the network, and we can tell when it’s delivered and alert, you know, alert the sender to the delivery of the mail. Sending direct ma- mail like that is, is um, it’s a new approach to the problem. It’s a lot more efficient, a lot more cost-effective, a lot faster. I mean, it used to be if you submitted a direct mail to a commercial printer—first of all, unless it was big enough, unless it was a couple thousand pieces, nobody would even look at it because nobody wanted to deal with it, yeah, um, but even if they did, you’re talking seven to 10 days just to get in the mail. And I don’t have any jobs that take seven to 10 days. I mean, we, we’re sending out very large jobs tomorrow or the next day, I mean, um… and the way we do that is with automation. We, we, we’re, we’re not touching any of those files in that way. I mean, we’re using technology to transform them into print-ready files. They can be printed, inserted, you know, use all the inserting equipment. We, we put all the marks on them and stuff to make the equipment work, um, so it’s produced and, and mailed very efficiently. And from, from the standpoint of our partners, if we’re handling the sales and the customer service and the automation, and all they have to do is produce and mail, then it makes life a lot easier for them, too. So, so they’re not worried if it’s 2,000 pieces, because the way I submit it, everything they get is going to get, be in larger volume and it’s going to be formatted correctly to work um, exactly the way their facility needs to, right? And historically, doing a piece something like I talked about with Speedy, with one piece of mailing, is just impossible unless you want to do it at home. Um, but in our case, we, we will batch that up with a bunch of other letters, you know? We lay down all the marks and stuff, make it a proper print ready file, and we’ll produce that through a commercial production facility. So it’s, you know, it’s better quality, it’s faster, it’s easier, it’s trackable, um, and I think it’s better for the environment. It’s just a win for everybody, right?
[00:14:52] Jeff: So I’ve, I’ve got a few questions. So number one, who’s your ideal customer profile—in other words—and then what sort of mailing are you doing, and then the third question is what sort of results are you seeing? Because the problem we have with digital overload and email is that it goes to trash, it goes to spam, open rates are just gone through the floor, um… so who’s your ideal customer profile? Let’s deal with that first. Who’s who is your target market that you’re providing the service for?
[00:15:28] Wilson: Okay, so what, so I think there’s a couple of different answers. So, um, it always varies for a second. I mean, because there’s lots of people who do mail. I mean, historically we’ve done a lot of work with realtors, mortgage brokers, um, independent insurance agents, but you can, you can see other people, you know, doing pest control, um, pool cleaners, landscapers—I mean, all kinds of small business, especially if they’ve got a targeted market to hit, um, this turns out to be a very effective tool. Yeah. That said, if I’m looking at our business overall, who are our ideal customers? Our ideal customers… we, we, we offer services to all of those customers, and because of automation, as I mentioned before, it doesn’t matter if it’s just 500 pieces, we’ll offer that service to everybody. But our ideal customer is somebody who is using the mail in a way that is strategic to their business, and they’re going to send jobs of 20,000 to 50,000 pieces two to three times a week. That profile fits debt collection, it fits call centers, it fits lots of different applications that, you know, that um, that might not be as familiar to America as general, but there are people doing that. And still do a lot of, we still do a lot of first-class mail, and you know, and when you think about mail, um, there actually, actually there’s a number of classes of mail, but the two most common are first-class and Standard A. Standard A used to be what we would call in the old days junk mail or bulk mail, right? And t- typically Standard A was associated with advertising mail. So if I’m getting something from Visa, Mastercard, or somebody like that, somebody sending millions of pieces of mail, I’m u-, you know, I’m using bulk mail. The reason for that is the bulk of the cost in direct mail is postage—postage about 70% of the cost—and in general Standard A is going to be about, you know, 20 cents cheaper, so there’s a, a delta. So if you’re mailing millions of pieces of mail, that makes a big difference, and so the people will tend to focus on that. If you’re doing smaller mailings, if you, if you’re like the, you know, the landscaper, the realtor, or whoever, then the difference between, you know, the, the savings is, is a, not as large, but b, you’re going to get a better class of service. So most of what we mail today is first-class mail, um, and most of what you’d see a large, a large commercial printer who’s not enabled service print would be Standard A, right?
[00:18:08] Jeff: Okay, so yeah, so um, in terms of you’re really using technology to make doing direct mail easier, faster, and cheaper—is that how you describe it?
[00:18:19] Wilson: I would, I’d say more cost-effective. I mean, um, you maybe, um, certainly in terms of um, the amount of time and effort involved, right? That’s what I’m talking about, yeah.
[00:18:38] Jeff: Yeah.
[00:18:38] Wilson: Yeah, one of the things that entrepreneurs, one of the mistakes entrepreneurs make, I believe, is they tend to look at their own time as being free. But, but time is one of those things you never have enough of—never—and you can never buy it, I’m sorry.
[00:18:50] Jeff: Yeah, exactly, so…
[00:18:50] Wilson: So you know, and, and if you look at the kind of thing we’re doing, I mean, I mean literally you launch a campaign in 60 seconds, you’re in the mail tomorrow, the next day you’re delivered, you know, anywhere in the US within three days usually with first-class mail, um… and you’re going to get, you know, feedback when it’s delivered, you know? So it, I mean, it is a very powerful tool. And I can explain why you might want to use direct mail rather than digital in some cases, too, yep, um, you know, if that’s a conversation you want to have.
[00:19:21] Jeff: So you, yeah, so okay, so that actually is one of the other questions I have, so… and how do you measure results is the other question I have, yeah. And, um, why should they use direct mail instead of, you know, email?
[00:19:42] Wilson: So, okay, no, that’s a great question, it’s an important question, I believe, yeah. Um, so I, I think that, you know, sometimes the, the discussion comes down to: why digital, why physical, you know, is, isn’t that stuff old, why would I use that? And, and yes, I, I mean it’s been around for a long time like some other mediums have. But I would tend to look at it more like a toolbox, right? As a marketer, you need a toolbox with lots of tools in it and you need to be able to use the right tool to solve the right problem. Yeah. And, you know, and so, and so that’s really where we go with this. I mean, if we, if we think about um, direct mail, um, the Direct Marketing Association says on average it returns $13 for every dollar invested. Um, so that’s assuming a well-targeted piece with a, with a well-performing offer, yes. So in general, if you’re using direct mail, the formula is, um, 70% list, 20% offer, 10% creative—whatever’s left is creative. Um, so we want to find the right audience, and if we can, if we can make the right offer to the right audience, then good things happen, right? And so that, that’s really the formula. So it turns out that the list is the most important part, not the creative. But most people focus on it the other way—they focus on the creative part, yeah.
[00:21:16] Jeff: Um, so that’s kind of an interesting dichotomy, yeah. You need, we need, you need to have people that are actually the right audience, otherwise it’s trying to sell you know basic… yeah, you don’t want to be selling something to someone that doesn’t like… okay, what are we want to sell? I want to sell fax machines to teenagers—of course, that’s not going to work, so…
[00:21:31] Wilson: Exactly, that’s why when you’re trying to boil the ocean, you know, you’re just trying to send to everybody, that’s not very effective. People can do that with tools like email marketing because the cost per send is, is very low, you know? If you’re doing a million emails, you may only pay a quarter penny a piece or something like that, you know, something very small, um… actually probably less than that. Um, but if you’re sending direct mail, every piece you send out needs postage, and postage costs money. So you’re going to have to, in order to produce a return, you’re going to need to make sure it’s very targeted and you have the right offer. The way to do that depends on the audience, and it requires testing. Um, so the, the, the question is, why, how do you measure? So the usually what we’re looking at is return, return on investment. So return on investment means dollars out versus dollars in. We have to make sure when we do that calculation, we have to factor in lifetime customer value—that’s a common, uh, omission, right? If I’m selling out, spending 10 bucks to get somebody to spend five, you say, “Why would I do that?” Well, except if he’s going to spend five bucks a month for the next 12 months, then that changes the equation, yeah, right? So direct mail, because it’s always been ROI, numbers-driven, that’s why it survives, because as soon as it stops making sense, people stop using it. But people are showing positive ROI, so they keep using it. And these days, we have lots of people who are new marketers who’ve only, you know, been exposed to digital that are rediscovering it, so that’s kind of an interesting phenomenon, too.
[00:23:15] Jeff: It’s almost becoming unique—in other words, um, if I sent you know a digital colleague I’ve never met before, but I sent him a gift and arrives in the mail, that’s memorable because everyone does email, but no one sends things via snail mail, so yes, it’s very memorable. The other question I have is, and talked about target, target audience, um, is there something you want to say about that, or…?
[00:23:43] Wilson: Well, no, I was, I’m going back to your other point, you asked me before, too… um, why I mean, you know, we’re talking about where this tool, what’s unique about this tool. If we think about a toolbox, and I can tell you a story about some ways we use it, but there’s a couple, couple things that are about mail that make it very unique. Um, first of all, it’s proactive, right? If I’m using Google AdWords or something like that, which is a great tool, fine tool too, we use it, um… but I have to wait for people to come find me. But there are sometimes when I just want people to find, I want to find them—I want them to respond to my offer. So mail allows us to be proactive—it’s proactive communication. Secondly, mail can reach everybody in the United States six days a week mostly, um… um, I’d say mostly mainly because some businesses are not open six days a week—some businesses are open, you know, five days a week—but to the extent that, that they’re open, um, I know how to reach them and, and I, and I can deliver a message to them. Um, if you’re talking about something like email, um, even though it has a cost, the cost advantage we talked about, being able to connect with everybody is not even possible—something like 35% of even legitimate emails never deliver, spam filters and what have you, um… and the third, the third thing is that, um, is that the targeting—the lists are very well-developed. So if I want to find Jewish dentists in New Jersey who purchased a car in the last, last six months, I can probably do that—I can probably get a mailing address. But trying to do that for a digital delivery point is really, really difficult, yeah.
[00:25:25] Jeff: Um, yeah, so on the, the other question I have too about this is um, is this more for targeting local, helping local businesses, um, or, or national organizations that are trying to reach local businesses, um… is that, I’m just trying to think about, you talked about things like dentists, you talked about, um, you know, florists, that sort of thing. So, um, is that the sort of like, you got the ideal customer profile, the one that wants to send it, but the audience they’re trying to reach is more local rather than national?
[00:26:01] Wilson: I, you know, it just depends on the, the, the customer and the application. Certainly, you know, if you’re a, you know, if you’re a florist, or a pool cleaner, or a realtor, then you’re probably focused in a local area, and we see a lot of that. But we do have very large national organizations, too—organizations that send to mail to people all over the United States, because with some of those businesses we’re talking about, you know, the thing that makes those potential customers a target is their location, right? They’re close to the place where we deliver a product or service. But there are some products or services that we might offer that are not location-specific. So for example, imagine I’m trying to do—this is not direct mail specifically—but imagine I try to do debt collection, imagine if I’m a mortgage broker trying to do cash-out refi. I mean, there’s no reason why I just need to do that in Oregon—I can offer that product nationwide, um, true. So there, so there can be a local component, but it’s certainly not a requirement, right? Um, the better requirement is, is that you know that I have a good list and a good offer, and, um, you know, I can generate a positive ROI, right?
[00:27:23] Jeff: So in terms of measuring a positive ROI, because, um, on the internet it’s basically you can measure it, but you sometimes trying to work out: how do people discover me? What was the first click? And the click could have been, it could have been an ad, it could have been an SEO, it could be… so how, whereas how do you actually measure that it’s actually the direct mail that’s produced it? Because that’s actually not digitally recorded. So how do you do that? Is that because that’s the only marketing they’re doing at the time and it’s coded or something? No…
[00:27:51] Wilson: Attribution, attribution is, I think what we’re talking about. Attribution is a big deal, you know, being able to figure out, you know, who’s responding to my offer, which ones are producing results and which ones are not. Typically with mail, we use a response device to do that—some kind of a response device. So, phone number—a certain phone number—fax number… well, not that people use fax anymore, but historically… specific landing page, um, okay, right? And, and we have the ability of our system, too, to generate a PIN, so we can generate a PIN number that’s specific to a client, to a, to a client and to a campaign, and that PIN number will identify where the piece came from. Historically, the biggest way that people identify that is by using a 800-number. So they use a toll-free number—you can buy toll-free numbers, there are lots of places to get toll-free numbers, and then you just program them to redirect to your office. But then at the end of the month, you can scan down your, your toll-free numbers and you can see which campaigns are producing calls, right? Okay, right. That makes total sense.
[00:29:07] Jeff: Because I used to do direct mail. I actually used fax for marketing and, um, it was a lot of fun. So Jay Abraham was one guy that I used to, uh, use as my inspiration for direct marketing back in the 1990s—he’s been around a long time. So all right, so you are generating data at the other end. You are using your tool to actually create, save you time and be very efficient to actually create the direct mail that’s sent out by your partners. So are you, to make sense of the noise and data—because you’re trying to find signal in the noise—are you using AI? And there’s going to be a segue from this um… are you using AI to help you make sense of the noise? In other words, what’s working, what’s not? Um, are you doing that at all, or that’s next, or we never…?
[00:29:49] Wilson: I, yeah, I would say we’re approaching it a little bit differently. I mean, it’s, um, so when we’re doing, when we’re doing network routing and we’re looking out at the environment, we have to take a lot of factors into consideration. Um, in today’s environment, everybody needs an AI story, and that would be ours, except I’m very realistic about it. When I talk to people, I say this is no, this is real intelligence—this is not artificial, you know? We’re looking at the factors in the network and we’re making decisions in an automated way, right?
[00:30:30] Jeff: So, so the other thing I want to ask, too, is okay, so what are the real numbers in terms of, uh, making it very efficient to send a mail—in other words, how do you reduce costs in terms of getting that all done? So what’s, what’s your numbers compared to other people doing direct mail versus you doing your technology to do direct mail? Is there any statistics or numbers in terms of “we save 20% of costs, uh, we deliver better, faster”? Um, what’s your data tell you about you versus your competitors, for example?
[00:30:59] Wilson: Yeah, that’s a good point. I, I mean, I, I’m not sure that I have, you know, specific numbers that we quote in that way. I will say that if you need to be in the mail, you, you have a piece of, uh, you know, mail job is 50,000 pieces and you need to be in the mail tomorrow or the next day, we’re the only people who are going to be able to do that for you.
[00:31:24] Jeff: Okay, well that’s the right answer. That’s what I was trying to find out—in terms of you make it easy and fast, would that be a good summation?
[00:31:28] Wilson: Yeah, easy, easy and fast. And we have the automation, so if we need to do other interesting things—you want to hook it to a CRM, or you want to hook it to an application, or, you know, any other way that you want to, um, play around with mail algorithmically—we, we have the ability to do that, because once we’ve captured it digitally, there’s lots of things we can do to it before we turn it into paper, right?
[00:31:57] Jeff: The other question I have, um, is that it sounds to me like you’re only operating from the USA, is that correct?
[00:31:57] Wilson: Um, we’re only producing mail in the USA today—the USA and Canada. And we, um, we had a facility, a dedicated facility in Canada, but we, we weren’t getting enough volume. But if we have the volume, that will return—I mean, that relationship still exists, okay? Um, I think this is one of the areas where I think Speedy has a really cool, really great role to play. Speedy, the application, they’ll even take a single letter. Speedy was designed to take the place of a postal meter—I’m not sure if you have those where you, you are…
[00:32:34] Jeff: Never heard of it, okay.
[00:32:34] Wilson: But here in the United States, a lot of large corporations have something called a postal meter. The leading provider in that industry forever has been a company called Pitney Bowes. Um, and the reason why um, they exist is because a lot of companies look at postage like money, and so they need an effective system to ma-, to manage this money. Like, somebody’s, somebody might, you know, take off with the postage or send their own personal letters, or what have you. So, so they tend to treat that postage as money. But, and the way that works is, some people print the mail internally in the internal mailroom, and then they run it through the Pitney Bowes machine. The Pitney Bowes machine puts postage on it, and then they still have to get it in the mail. But from our perspective, that’s really not the biggest part of the problem, right? I mean, so I mean, I know how to get stamps and I know how to get postage. Uh, the really hard part is to, you know, is, is that I have to create a letter, I have to print it, I have to fold it, I have to insert it, I have to address it—I have to address it properly—I have to get postage, and then I still have to get it in the mail, yep. Which, if you’re a mailroom, means you still got to put it in a truck and drive it somewhere, or somebody’s got to, can pick it up. So there’s a lot of pieces missing. With a tool like Speedy, you can just create the letter, push the button, and boom, it’s gone—you just, just bypass that entire process.
[00:34:04] Jeff: Wow, and that’s here…
[00:34:04] Wilson: That’s probably a $6 billion market—postal meters, wow. But because the way the tool is designed, consumers can use it in the same way. A consumer can put a letter, because we don’t care whether it’s a business or a consumer. And if, if you have your consumer who’s in another part of the world—you know, wherever that might be—um, you know, if you’re in New Zealand, you could drop your letter on, on this icon, it will, you know, it will fly across the ocean electronically, end up in a queue here, be, be, you know, be batched, processed, printed, folded, inserted, mailed next day, and at a fraction of the postage. You don’t have to pay international postage. Oh, and by the way, nobody had to pay for any jet fuel to get it there.
[00:34:55] Jeff: Okay, so you’re distributing the distribution, yeah.
[00:35:00] Wilson: So, so if, if you’re doing international mail, if you’re in an international location and you’re doing mail to the United States, there is no more efficient way to do that.
[00:35:08] Jeff: Very cool. All right, I’m going to segue to something very, very non-direct mail, because you’re in the middle of doing a degree, I believe, at Harvard University, or have done it, on… which has a lot of AI in it, I’ve interested in, um… and you’ve been in business for a long time, you’ve done six startups, you’ve done raise venture capital, and it sounds like your brain’s running at a million miles an hour, um, all the time. So you, you’re studying AI—what are some of the things that you, fascinated you about AI, and a generative AI, for example? And what are you, what are you learning? What are your insights from your Harvard experience, education experience about AI?
[00:35:48] Wilson: You know, AI is interesting, but you know, the AI that we see today, it’s a lot different than… well, let me, let me back up and say, even before the PhD program, right? I started as a software engineer. And even back in the ’80s when I started as a software engineer, people were talking about AI, except AI looked more like, you know, like HAL from 2001: A Space Odyssey, you know, the movie. Just stop in…
[00:36:15] Jeff: Yep, exactly.
[00:36:15] Wilson: So, uh, along the way people have tried inventing languages like Lisp and others, they’ve tried um, creating knowledge bases, they’ve tried, um, expert systems, you know? There’s, there’s been a lot of different incarnations of AI along the way. In, in this particular case, we’re looking at large language models, and they’re really looking at pattern matching. And based on patterns that occur, um, in data—different pools of data—then being able to generate new ideas based on prompt. It’s a completely different… I mean, artificial intelligence for me is more like—have you ever seen the movie I, Robot? Yes. Will Smith. Yeah, I mean, that looks like artificial intelligence to me, but, but this is a step along the way. So I mean, all of that stuff has to evolve. Um, because we’re doing pattern matching, there are some natural limitations built into it, like there are lots of examples of, you know, people doing 1 + 1 = 6. When he said, well, how’s that work? Well, because AI is not really doing the math, it’s just looking at an image of a one and an image of a one, and it sees enough images of one that match it, you know, that are matched with an image of six, it just figures that’s the way it should work, you know? Um, the other thing I will say is, I, you know, the, the place I’ve spent most of my research time most recently is on innovation—you know, diffusion of innovations and stages in the life cycle. I actually did a PhD dissertation on it, so that information is public—you can find it in, um, Google Scholar. Um, but basically, depending on the point of the life cycle, different kinds of innovation are different, more important—whether that’s product innovation, process innovation, marketing innovation, organizational innovation. Um, and when we’re talking about creativity and innovation, that brings always brings me back to generative AI. And I think that there’s some natural limitations—and I’m not saying these won’t be overcome, I mean people definitely will work on them—but if you think about the way people, the way AI would approach that with a large language model, is they look at all the successful outcomes, and based on these successful outcomes, um, or popular outcomes, they’re going to forecast a new one. But that’s not really what, the way the innovation happens. Innovation is a process of trial and error—we learn as much from every mistake as we do from every successful venture, sometimes more from every mistake, you know? They asked, uh, Edison one time, he said, you know, “You failed 100 times making a light bulb, why do you keep doing that?” And he said, “Because every time I try, I learn, yeah, there’s one more way I, I’m not supposed to do it. And process of limitation, it gets me closer to a solution.” Yeah, um, but AI, the way it’s conceived of today, doesn’t really go through that step. So there’s a whole bunch of… I think without having access to the mathematics behind it, um, there are a whole bunch of false steps or mistakes that we don’t get to see. And because we don’t see them, we can’t take off from them or evolve from them, right? Yep, okay. And then the second thing is, be-, because of the way AI is designed, it can never know when it’s finished, right? I mean, it, it doesn’t… the only way it can, it can be finished is if a human says, “Oh, that works, that makes sense,” right? So, so it, not only can I not make the mistakes, but it doesn’t… that process of going through the iterations to arrive at a fin-, final product is problematic, too. Because I don’t have some, you know, some being there to say, “Oh no, that’s right, oh no, that’s wrong, take a right, take left, take, you know…” Have you ever done, um, brainstorming, you know, with a, with a group of people, and you know, you’re in a room and the one thing you say, you don’t say, “That’s a bad idea,” you just try to capture all the ideas you can say, and then one person will say something and somebody else will say, “Oh yeah,” they’ll play off that and play off that and play off, you know… and, and that process is what leads to new innovations and new creations. But I think with the technology we have today, it’s difficult to get to that um, impossible, it’s just trickier, yeah. Um, so when we, we can use um… and most ideas, I mean, I mean, thinking out loud, most ideas are not really all seminal genius, right? All some new idea. A lot of times it’s a, a combination of old stuff, maybe even unrelated old stuff that we put together, you know? Mail and the internet. So, and sometimes the alchemy works, and other times it doesn’t. Exactly, yeah. I mean, I mean, if, if every pattern in the future, if every solution in the future is going to look like all the patterns in the past, then, then that’s not very interesting, is it? Well, there, well, there’s no future then, there’s no innovation, yeah, yeah.
[00:41:39] Jeff: Okay, interesting. All right. Yeah, so I, I think, I think it’s really cool and it’s exciting. People are doing some great things, certainly generative AI has a lot to bring. Um, it’s not the Holy Grail of AI, we’ve got a lot more to do.
[00:41:51] Wilson: Oh, yeah, we just started. We’re essentially two years in the democratization of AI, which is ChatGPT and Chat.
[00:41:57] Jeff: So, uh, what’s happened… I am blown away by the velocity of change and development, um… and we’re seeing basically about a trillion dollars in data centers, you know, and investment in the large language models and so on. So the amount of money pouring in is going to cause a bubble, but again, we are laying the foundation for continuing to evolve AI to be our assistant, to be our guide, to be our coach. Because I think for me, with AI, I, I’m watching a game. All of us come from a fixed past, brought up a certain way—went to the same church, went to the same, um, you know, went to the same, um, you group, football, college, whatever it is, right? So we’re… and we got, we go to college, so we got, we’re all raised in bubbles, aren’t they? Different bubbles. So we end up having quite fixed thinking. And what I love about brainstorming with AI is that it can come up with all combinations you haven’t even thought of, thought of. And that’s for me is that, I think the AI can break our own vertical bubbles that trap us in the past, but we need to be open to the future to find the innovation and recombination of the world’s intelligence, which is what’s sitting in knowledge—which is sitting in the cloud, um, in ones and zeros. Um, so we actually, for me, I find it exciting because I ask a question of AI, and it’s tapping the world’s intelligence, the world’s wisdom, the world’s, you know, knowledge, and it comes up in ways that I wouldn’t have thought of. And then I can recombine it myself as well on top of that. But it breaks a bubble of our templated thinking that traps us almost in a prison of our own mind. Right? I mean, you think about some of the work with, um, companies like IDEO and others have done, right? The, the idea is that you’ve, you know, you have, if you have groups with, that have diverse backgrounds and life experiences…
[00:44:01] Wilson: Much more creative, yeah.
[00:44:01] Jeff: …then you can get better solutions. I mean, if you think about a Venn diagram and, you know, there’s one person, they have certain amount of knowledge and life experiences. If we have a lot of different people with non-overlapping, you know, spaces, then our, our solution set, potential solution set becomes much larger, exponentially larger.
[00:44:25] Wilson: Yep, that’s what I’m saying.
[00:44:25] Jeff: And that’s what makes it exciting, um… and I think for me, I see AI as both an amplifier and accelerator of human intelligence, um, as our, you know, as a coach assistant that has almost perfect knowledge, um, that can recall it in seconds. So for me… but it doesn’t mean that we hand over the keys of the kingdom to the AI—we actually use it as our coach and assistant to help us, um, break boundaries down that we quite often as humans are locked in, and that includes bias, um, includes, you know, racial, gender, whatever it is. But what it does, I think it, it opens our mind to a whole world rather than be trapped in our own world, right?
[00:45:16] Wilson: No, that’s true. I mean, the thing about AI is that from an algorithmic standpoint, it’s data-driven, so we’re looking at the data and saying, “What does the data tell us?” But data can have bias.
[00:45:26] Jeff: Oh, sure.
[00:45:26] Wilson: You know, so if, if, you know, if, if you’re, you’re tapping into the articles from Fox News, you’re going to get a different result than if you’re tapping into the articles from MSNBC, right? I mean, there’s, there’s bias in that data. Yep. And, and the language model can’t tell, it just, it’s just looking at data and aggregating it in the best way it can, yeah. And the other thing I’ve come across recently is the idea of Sovereign AI, um… because the trouble with capturing all the world’s data and putting it into one bowl of soup is that it all becomes homogeneous. And the problem with that is that then the world all starts to look the same, which is boring as bat shit—I don’t want the world to look the same, I go to Europe to see Europe, not to see Australia. Um, so for me, I think human imperfection brings its own joy as well and, um, you know, perfection’s overrated, because job done, might as well fall asleep, right? But…
[00:46:24] Wilson: Right. I mean, I mean, with some things, it’s the imperfection that makes it interesting. I mean, like, I mean, think about art, you know? Monet, Degas, things like that. I mean, you could take a photograph of those images, but what makes it interesting when the artist does is to get, you know, their own take on it—they, you see it the way they see it with all its imperfection, yeah.
[00:46:53] Jeff: Yeah, yeah, yeah, it’s that for me, too. He’s done some blind testing with even poetry, and actually, um, AI—in other words, poetry created with AI—they’ve done blind tests, people viewing it and say, “Tell me which one you like best.” And the scary thing is that most people choose the AI-generated art and poetry.
[00:47:15] Wilson: Oh, yeah.
[00:47:15] Jeff: Which is a bit scary, but on the other hand it’s rather interesting. I read Yuval Harari, um, you know, the guy that wrote Sapiens, um… he’s just written a book recently which he’s worried about AI, because generative AI is starting to tell stories. And what makes us human is our stories. So, and when I, you go to write an introduction, you know, humans are, humans are pain and happiness machines—we actually learn from pain and… well, we don’t, um… we’re all trying to find our own story amongst all that noise. So for me, his worry is about the fact that, um, AI is going to start creating stories that we’ll believe and that will change humanity. So it’s quite a fascinating take by him because he’s very much into humans’ uniqueness is our stories.
[00:48:14] Wilson: Yeah, the only thing that it makes me think about is the fact that a lot of the great literature and great art, and, and music too, I mean, is driven by emotion, right? It’s, which is, which is your story that you’re telling, whether it’s in whatever media it is. And, and emotion is not data, I mean, emotion is something different than that. So I, I mean, I don’t think… I mean, you might be able to, to create a table that says, “See this, generate this emotion; see this, generate this,” but it’s still not going to be authentic emotion that you’re trying to generate for a reader or a viewer or somebody who’s engaged with your art. Um, I think that’s really tricky. I, I, I think you can do replication. But, you know, one of the things about music, I mean, I, I think about, too, in every generation you’ll have great bands, you know? We listen to rock music, and some of that music, it’s not all just good music—some of it is angry, right? You get, get this anger, or, you know, or, or, or lovesickness, or whatever, whatever that emotion is, and they manage to turn it into this powerful music that communicates that to somebody else. But that’s not just about data, I mean, I mean, there’s something more there.
[00:49:40] Jeff: Yeah, a lived experience. Exactly. So for me, I’m fascinated by the intersection of human and machine, and I think machines actually asking us big questions like: what does it truly mean to be human?
[00:49:52] Wilson: That’s a good point, yeah.
[00:49:52] Jeff: So, I don’t have the answer.
[00:50:01] Wilson: No, well, that’s, that’s what, that’s the fun—is the journey, not, not arriving, um, right?
[00:50:09] Jeff: Yeah, and, uh, so for me, too, the hom-, you know, the homogeneous nature of AI is that it, it’s basically not allowed to do bad things. So what sort of try to, try and program it so it only gives positive things, so AI can’t sound angry, otherwise it’s turned off.
[00:50:24] Wilson: Well, if you… but if you want to capture emotion, you have to do all emotion—you have to do anger, you know, anger, joy, sadness, all…
[00:50:33] Jeff: Exactly, so yeah. All right, anyway, I’m fascinated, um, and, um, I’m having fun exploring and writing about it. I, I write to learn, I read to get ideas and information, but then I got to make sense of it. So, and part of trying to make sense of it is actually having interesting conversations and hearing stories from smart, intelligent people around the world such as yourself. So, um, thank you, I appreciate that.
[00:51:02] Wilson: Yeah, it’s, um, this brings me joy.
[00:51:02] Jeff: So that leads me, and a segue to: so for you, Wilson, if you had all the money in the world, uh, what would you do every day that would bring you deep joy and satisfaction?
[00:51:18] Wilson: That’s a great question. I mean, it’s, it’s an important question, too. I mean, certainly one of the ways we measure success, we keep scores with money, but at least in entrepreneurship, um, that shouldn’t be the goal and that shouldn’t be the primary goal. And the reason why—and don’t let investors ever hear you say that!—but, but the reason why is because, you know, honestly, most startups fail. I mean, they’re, they don’t, you know… I think it’s nine out, I think it’s nine out of 10, or even exactly… a lot of entrepreneurs in order to do a number of startups before they get one that gets traction is successful. So, so and I’m not saying that to say you shouldn’t do it, I’m saying there has to be more in it—there has to be some other passion you’re pursuing or some other need you’re fulfilling. And if you can make a lot of money along the way, that’s a great thing, too. Um, so for me, I, I really like to build things. I mean, I like to start, start with an idea—start with an idea that I think could change the world and then try to make it real, okay? And if I can do that, that feels really good for me. And so there’s the, the initial idea, there’s additional things that work on it—we’re solving problems. And in my case, because I can’t paint or sing or… I’m using technology to do that, you know? Um, so I enjoy that a lot. I, and I like, I like the idea of making a difference in people’s liv-, lives, making a tangible difference and making the world a better place. And if I can do that using my art, which is technology—my art form which is technology…
[00:53:01] Jeff: That’s your tools, yep.
[00:53:01] Wilson: …that’s a really good thing for me, yeah.
[00:53:01] Jeff: So you are doing what we designed as humans to be, and, uh, we are creators and we are also… watching over the millennia we’ve seen that we are actually creative tools to help us do things better and faster and make a difference. And, um, you just said that what brings you joy is creating and I love that. And creating, and, and sometimes understanding, too—looking at a complex problem and just saying, “I wonder why it works that way, let’s find out.” And that’s the other part of being human is, um, some have more of it than others, and that is curiosity, yeah.
[00:53:47] Wilson: Yeah, yeah.
[00:53:47] Jeff: Yeah, for me, Joseph Campbell, um… for me, uh, which brings me joy and part of my creation is writing, um, nice, and communicating, and listening, um, and trying to be better at it every day. Um, as Joseph Campbell says, we, we, there is no essential purpose to our lives as humans except to follow your bliss. And, um, for me that is, and follow your bliss means you’re following your curiosity—what gives you joy, what, what sits inside you that says, “When I do this, this fans great,” like you were saying you’re creating from technology tools. And for me, I am curious about something. And what I do now is, and I’ve done it for the last 15 years now since I started my blog on, because I was curious about social media, was to actually… I follow my bliss now by curiosity, but then create out of that curiosity, that’s… so I think as following your bliss and then actually creating something out of that is, I think that makes me as human as I possibly can be. And, um, and then paths will open and doors will open, um, that you never expected. Because we all think we got control—at the end of the day, most of the time we don’t.
[00:55:13] Wilson: No, no. I, I actually… I mean, I, I think those were good, great points. And I talked with an entrepreneur one time—it was a gentleman who owned, he started a window company, very successful, ended up selling it to Viking Industries or something, it was um… made millions of dollars, hundreds of millions of dollars. And I asked him, I said, “What is your secret? I mean, if you had to tell an entrepreneur what’s the secret to being successful, what would you say?” And he said, he said there are really two elements that you need to think about, think about, spend some time thinking about. One is: think about your passions—everybody has stuff um, that they feel passionate about, feel strongly about. And he said, and then think about your gifts—the things you’re really good at doing. Everybody has some gifts as well. And the right opportunity is probably the one that lies right at the intersection of those two things: something that you’re really good at but that also brings you joy.
[00:56:14] Jeff: Yeah, that is a great point. And there is a fantastic model, uh, for success, for happiness, and it’s called the Ikigai model, which is out of, um, Japanese philosophy, which is the intersection of those. And you just mentioned the word intersection of those two, but check out Ikigai. It’s really fascinating about what are you good at, what do you love doing, also what’s your experience, okay? And then what will the world pay you for, okay? So have a look at the Ikigai model—it’s fascinating, I think. And for me, it’s the best summation of, of, uh, what you should be doing in life sits in the intersection of all those questions.
[00:57:02] Wilson: No, I’ll do it. It sounds great, yeah. I’d love to, yeah, I’d love to explore more.

