Brad Kugler on Why Not Doing AI Is More Dangerous Than Doing It

Hello and welcome back to Print Island.

Here we are on the software matrix today and we are back talking about all things AI pavilion and printing AI for Printing

United 2026 in Las Vegas. Now I have a very special guest, somebody who was

part of the inaugural AI pavilion at Print United in Florida, not too far from where he is now, I’m assuming. Brad

from Direct Mail 2.0. Welcome to Print Island.

Thank you, Chris. Good to be here. Happy to talk about uh any experience you want to talk about.

Well, in that case, let’s find out a little bit more about what Direct Mail 2.0 is, what it does for the print industry, and and let’s kick things off.

Great. I think the easiest way to describe it is we are the tech stack to enhance and improve direct mail. So we

have a couple of brands uh first being the direct mail 2.0 which uh reports enhances and improves the engagement on

direct mail through a bunch of technologies. There’s 15 of them on the platform. Then we have who’s mailing what which is a competitive intelligence

database that allows people to kind of see what their competitors are mailing and and who to who and and what they’re

mailing about. And then the most recent platform is dm2.ai which is a propensity modeling database.

So, I’ve got tens of thousands of direct mail campaigns and we’ve spent the last 18 months building our own algorithms

and integrating with with Chat GPT and Claude and Gemini to basically predict

or help a person design and deliver the best possible mail campaign based on the past results of about 80,000 campaigns

we have information on. So, we like to say that we are the tech stack for improving direct mail.

Okay, that’s quite an introduction, sir.

I’m going to focus in on one of those elements there. I’m going to talk to you about direct mail to.ai.

Have I got it right? Y.

Now, you mentioned to me that you launched this in beta last year at the inaugural uh AI pavilion. Um, and

officially it came out last month. Is that is that correct?

uh the end of March we we came out of beta meaning the freebies were gone and you know that that beta period we

learned a lot you know I I I have new respect for those large language models and trying to build something that

nobody’s ever built uh obviously I’m I’m bootstrapped there have access to

billions of dollars uh I would say that I didn’t even spend a million dollars on what we put together a and considering

ing the scale and I guess the unscale that I I’m pretty impressed with what we built. I mean we AI is built on data and

we have a pretty good data set that gets better all the time.

I was just about to ask you actually what’s the data behind it.

I can go into that a little more. So So it really comes from our other two platforms. Direct mail 2.0 know which we

have all the engagement data on almost 80,000 different direct mail campaigns, two billion pieces of mail going back

seven or eight years. And then we have the company we acquired in 2024, which was who’s mailing what, which has over,

I believe it’s 40,000 advertisers that have sent about half a million direct mail campaigns. So, we’re leveraging

that massive payload of data and attribution and build models from them based on what people are currently

mailing. I I don’t think there’s anybody that’s got the direct mail data that we have.

Okay. And that’s that’s quite a statement as well. Um, we may have all seen yesterday that Microsoft stopped a

large proportion of their team using AI tools due to the cost. Um, what’s the pricing model? Uh, and

so so well, we have two models as everyone else. We have we have a one-off model where you can just use it one time and never have to pay us again. And if

somebody’s got a campaign, they just want to check one thing. It’s anywhere from $19 to $99

depending on the amount of AI you want to use. And and let me expound on the $99 because that’s really where the

value is. uh we have our own proprietary model that uses our own database and it’s been handmodeled then we once we

have those results we go out to not only chat GPT and Gemini but then we use a third we use claude to sort of look at

all three of the recommendations and models string them together throw out the bad improve the good so we call it a

four model analysis all right so obviously if I’m using all of the top frontier your models as well as my own model I have to charge for

and we even allow one better now that this is version 11 one that rolled out last week we allow a retest so you get all these recommendations

to improve your mail piece and then let’s say somebody acts on they want to re-upload it or they want to change a couple of parameters because they’ve now

decided to make some changes we allow that done for free so essentially you get two tests for $100 no commitment

never have to do it. Pay us another dime. And then of course there’s various subscription models that allow you to use it in maths. And based on your

appetite, it goes anywhere from, you know, $79 a month all the way up to $390

a month on a subscription basis. And that obviously gives you, I don’t want to say all you can eat, but probably more than you would need. It

would be more for people that are commercial printers or they’re agencies that have multiple clients. They’re not a they’re not a landscaping company that

does a spring cleanup and they don’t need to run it more than once or twice a year. So, we have we have plans for all budgets.

Okay. So, um you built something generally quite different by running your ma mathematical model through

Gemini chat GPTO then triangulating the outputs. Why three engines rather than one? And what

does that give your customer that a single LMM tool cannot solve?

That’s a great question. Okay. So, so this is what we determined in beta. And I I’ll be I’ll be brutally honest. So,

we we spent hundreds of thousands of dollars hiring, you know, these math geeks that build algorithms. All right?

and and algorithms are just a series of mathematical computations based on certain benchmarks and where do things

line up across an average or a median and then based on what that is you look at attributes that were different that may have lifted or dropped you below a

certain benchmark. So it was very mathematical it was very dry and it was only related to that specific piece in

that specific industry. So, the results that it gave in the first

beta pass were a bit generic. I’ll be honest with you. Make a bigger headline, you know, use brighter pictures, you

know, it was it was it was things that people were like, I I don’t need to pay for that, you know, and and how much would it be different

than using chat GPT? They were very generalized. So, we needed the the magic

of both. We needed the massive computational power of those LLMs and we needed our own proprietary data to get

the right output. So we would do our calculations, run it through one, run it through another, then have the third compare and contrast and give the final

report. I think it’s the it’s the ultimate because what are we doing?

We’re saving people the time of going to chat GPT and they may think why do I need you at all? I’ll upload my creative

design into chat GPT and it’ll give me a response. Yes, it will. It’s going to give you a generalized response based on

best design practices. Maybe even not for direct mail, just in general. It’s not going to have data for specific

industries, for specific time periods and specific geography. Okay? So, you can do that. You are not going to get the specifics that a professional needs.

What is it that nonprofits are doing at this time of the year that’s successful?

We drill down into the nitty-gritty of that. And not only that, we’ll show them competitors that were in their category

at the same time. This is where I’m leveraging my who’s mailing what database. So, we’ll take their campaign and now we’ll draw off of our who’s

mailing what database and show them a couple of pieces that are competitive to them. So, so not only do they have our recommendations, they see what their

competitors are doing real time. Let me ask you, you think Chat GPT or Claude has that data? Absolutely not. And they never will.

So, this is where we add enough value that that those companies will never have and never be able to bring.

Um, fantastic. Obviously, we keep our stuff in in a silo so they don’t then get access to it. uh we’ll show them the singular piece

but nothing beyond the nothing beyond that not the data behind it that gave us the algorithm for the output so

wow that’s really insightful yeah really interesting um I didn’t actually expect you to unpack it with that much of a USP

either before we move on to why we’re here today which is the AI pavilion I’m going to ask a few personal questions if you don’t mind um you seem pretty up to

speed on on you know the industry and where we’re going. But it’s very niche part of of our industry. How comes this

part of the industry? If I could start off with that first personal question.

Um, I’ll tell you why like why did I pick propensity modeling for AI? Very simple. I I’ve been running and managing

direct mail campaigns for nine years and I kept all the data. All right. And again, we’re not sharing by individual.

no PII. We’re sharing by vertical and by zip code. So, so our customers that have used our platform to help monitor the

engagement of their campaigns. We’re not sharing who they are. We’re not even sharing who the companies are. So, I’ve

accumulated this data what I feel is probably one of the best attribution

mail that’s ever been built. And I knew that there this was monetizable.

It took the advent of AI a couple of years ago to be able to look at the creative, look at the data, correlate

the two together, and find what happened with the creative or targeting that lifted the engagement on the other side. Now, to clarify, I do not have ROI data.

I have engagement data. Who called? Who went to the website? not who but how many people what kinds of people uh what

people clicked on links um those type of things. So we know if that creative is getting traction. So to me it was a

natural and it really took getting the technology to a point where it was ease easy to build this and with the advent

of the LLMs and the explosion of AI tech it was able to build this and and honestly nobody’s done it and I’ve got the best database for it so it’s got to

be me you know it’s got to be okay again a very strong statement we love that so let’s keeping up with the personal theme here uh previous business

of 23 is uh I would call you a disruptor. Is that a fair a fair statement? But um I like to call you a positive disruptor.

So So a little history. So So my previous business I was in uh distribution of packaged media that people know that as CDs, DVDs, Blu-rays.

All right. So uh probably nobody out there is still buying those products now. Uh, I was one of the first people

to start selling those online via eBay and Amazon in the late 90s. I got the the opportunity to meet Jeff Bezos a

couple of times. We were originally one of the top sellers on Amazon in the late 90s, early 2000s. Um, it was a great

business and we knew disruption was coming with streaming. It was this was not anything that was not foreseen and we had tried

many things but I knew that I was going to get get the smackdown and I did and

in 200 we peaked in 2012 and by 2017 I watched our company go from 25 million

down to under a million dollars. I I wrote it to zero essentially. And before I chose my next business, I wanted to

get into something that was a disruptor, not disrupty.

And direct mail, digital technology, and tracking technology sort of fit the

bill. Now, there’s been pressure on direct mail, and it it’s it’s held its own. You know, it hasn’t crashed, per se, and these people direct mail’s dead.

Direct mail is dead. probably the only reason it’s still alive is because people like us that have taken this technology to the next level. I’m not

saying just me. The the inkjet printers, the variable data, the the automation has has allowed economies of scale to

keep this industry relevant. And not only that, yes, it’s very expensive to deliver content to targets dollar or so

a piece, but it is one of the most effective marketing technologies. So if we can come along with new technologies

and make it more effective and better then this industry survives because of that and again it’s the hardware guys

it’s the marketing guys it’s it’s the inks it’s the textures it’s everything you know there’s a lot of people that watch this

channel would absolutely agree with you on that sticking with the personal theme um so into the next stage of your life you’ve

done all this with three kids am I right in saying that you got three Yeah, I would say that my kids have have have grown. Um, it was unfortunate when I was

50 years old and my previous 23-year-old company was going down the drain. I had three kids in college or about to be and

that was scary to start over at 50 years old.

Um, that that was a lot of sleepless nights. You know, you’re burning cash.

You have an investor who’s getting nervous after a year like, “Hey man, when are we going to stop the bleed here?” you know,

and uh that it was tough, but I knew that I really didn’t have a choice. I had to get back on the hustle train and

build something again. You know, I had gotten comfortable and coasted and then I wrote it down and now I got to build a

company from scratch at 50. And I’ll tell you, building companies as a middle-aged or 50-y old guy is not

the same as it was when I did it when I was 24. Okay. You know, the 60-hour weeks,

no kids, it it was it was a different experience, much easier than it was doing it at 50, I’ll tell you.

And do you think also that’s a really valid point and I think um there’s a bit of an entrepreneurial message in what you’re saying here. But do you also

think that, you know, 20 30 years ago when you were doing it that the industry, the economy, the global economy has collapsed twice, isn’t it?

Three times. If you take in 2000, I’ve now been through three three, if you call COVID, four recessions.

And I’ll tell you what I’ve learned having been in business trying to earn a living through four recessions and I’m I’m considering now entering one probably as well.

You learn that the sky is not falling.

There is an there is an end to this and it’s more about perseverance.

There’s some hunkering down, but as my old Jewish grandmother would always say, “And this too shall pass.” Okay, after

seeing it three times, it’s just a matter of of make getting through it and there are more prosperous times ahead. I

mean, when I went to university, we talked about the 8 to 10 year business cycle. This was 40ome years ago. It is proven true. These things come and go

every 8 to 10 years. And we’ll get through this one. And then we got another one eight to 10 years down the road. So get ready. I’ll probably be

retired hopefully for that one. Won’t be my problem. You know, I think that’s some really good advice there, Brad. And I think um that applies to everybody in this industry right now.

So I I think we really appreciate that.

I’m going to I’m going to switch it a little bit back to um the pavilion now.

Um you’re one of the founding participants last year in in uh Florida.

A business. What did it mean to you guys? um being there from the start and

um you know was it worth the bet? Was it worth the investment? You know, like the gamble on on that first pavilion.

Absolutely. I mean, we were launching an AI product that had never existed before in an industry that may be slower than most to adopt these type of things.

And of course, it was a gamble, but it to me it was one of these things. Well, I’m going to release a whole new brand,

a whole new AI platform right at this time and I’m not going to be in this pavilion. I mean, it’s more of what I

stand much more to lose than to gain. I know exactly what I would lose if it didn’t work out. But if I didn’t do it,

who knows? The loss could go much deeper, you know?

So, anyways, it was it was a no-brainer and I feel that way again. uh in the year or six or 10 months since we have

all of our platforms are AI first in terms of architecture. We have retoled everything

in an AI sort of architecture whether we bolted on new services to our existing legacy platforms or we’ve just retoled

it from scratch. So I I will not be disrupted again. So, you know, this was

this was a no-brainer and obviously I’m going to be there again this year and

I made a post on LinkedIn the other day and and you know 20 years ago as the internet was booming and I wasn’t in the

printing industry but I was in a different industry and then there was there was internet pavilions at 25 years ago.

uh this is going to become so prevalent that you know I’m not trying to discount what printing has done they’re focusing on this but in five years everything

will be AI and there will be no AI pavilion you know if you don’t have AI you are a dinosaur and you will be

disrupted and you will be replaced all right and and here is the analogy imagine

25 years ago you’re entering the workforce and you have no

experience or knowledge of any of the Microsoft Office tools. All right. How far are you going to get in the business

world entering 25 years ago with no experience in email or Outlook, PowerPoint, Word, spreadsheets? This is

the precipice we’re at now. If you are not using AI tools on a daily in your

life, you are the same dinosaur of the guy who never learned office tools 25

years ago. So now there’s there’s a a group of us that are retiring out where maybe it doesn’t

matter. You know, they’re five years they’ll they’ll be in their rocking chair or being a consult. I don’t know. But for the people coming

up, if they are not on claude or chat or using one of these tools and not just on them and not just asking questions to

make a better email, but building a gentic applications to improve your daily workflow.

Yeah, you’re going to have you’re at risk.

Let’s just put it that way. You are at risk in this market three, four years from now.

It’s interesting. Uh I was at an event last week in Europe and I really felt as if the tide had changed that there was

generally um people that were uh would never touch software were now vibe coding their own products and actually

the requirement of where software will be purchased and used now fits in with an AI product of some sort and if it

doesn’t then it’s it’s it’s going to go extinct like a dinosaur, right? Um, very very interesting and it’s very interesting to hear you say that very

strongly that you know AI pavilion is here now to help us get ready for the big change that’s coming. Um, when you

at the pavilion last year and I think one of the things that Print United are trying to do is is like you say prepare the industry for what’s coming next.

What was the reaction from the print shop owners to to you know what you were showing them? What were you surprised about the most from what you heard?

Because let’s let’s remember Brad, the sessions were all packed, weren’t they?

The the the pavilion, especially Adobe, there must have been 500 people crowded around. Yeah. Yeah.

You know, um they all carried the same theme. This is here. It’s not going away. The toothpaste is out of the tube.

You know, that was and to mention this is the worst that AI will ever be today. Every day it improves and gets better.

So, in regards to the conversations I had and the and the things that I was showing people all week at the show,

everyone was very impressed and and again, that was our worst scenario. But then you have to everybody’s very

validative and very impressed and they love the technology, but you know, there’s still a friction point to being able to where they got to pay for it,

you know. Uh, yes, a lot of people have gotten on that $20 a month bandwagon for the pro version of chat or claude

because it’s easy. You you you get a better email, you write a better letter that it’ll fix your spreadsheet just

telling you what to do. Okay. Um, that was the start of it. Um,

businesses are adopting it with with some trepidation, especially in a hard economy. You know, the thought of going

and spending money to implement these things is scary because there is no guaranteed ROI. You

know, this is uncharted territory. And and I understand if I I’m a business and I’ve got to invest 10 or $20,000 into AI retooling.

Uh that’s scary because right now maybe my my numbers are down and I don’t have the 10 to 20 grand. So what do I do? You

know, and and it it is a tough call, but my thought would be to not do it is more dangerous than to risk doing it now.

Interesting.

Even if you don’t have the money, honestly.

So I was going to say I was actually going to ask you for some advice for what you’d say to uh printers, but actually that was pretty strong really

is lean in and and start talking to people like yourselves, I guess. Um, but in your opinion, um, why should PSPs

print business owners make the trip to AI Pavilion again in 2026?

And I guess when I say again, let’s also say in general, why should they come?

What’s different this year compared to last year?

At the very least, if they’re not ready to make an investment, they need to open theelves up to the operational experience of what these technologies

can do and at least start thinking with them in the future. So, I don’t care if you don’t have the budget for it and you’re not going to implement some AI

workflow and you’re not going to change software, you’re not going to buy something. At least expose yourself to the possibilities because it will get

you thinking to how this can help make your business more efficient and and and and I am by no means someone who’s

advocating that, you know, do this and you’ll be profitable or you’ll operate with one half the staff. I I’m I’m not

promising that, but you’ll be able to see ways to kind of navigate what I believe are going to be tough times the

next year or two. And when the time comes, at least you’ll have gotten over the educational or

operational understanding of what these tools can do. and you’re not starting at square zero because it’s easier to keep up with it than kind of start fresh in a

year or two knowing nothing when everybody that’s going to be talking to you is already going to be two or three years deep in this stuff and you’re you’re coming out of the weeds. So begin paying attention at the very least.

Okay, that’s interesting. So um your hot take, what’s the biggest misconception in our space right now with AI?

Quite a loaded question.

you know, it it it goes back to those that feel that it’s going to be the ruin of everything.

Um or it’s, you know, it’s the devil incarnate. You know, uh I

don’t know if you follow the news, but you know, our our gracious Pope Leo wrote one of those reports saying that this is this is not good and everyone

must keep their eyes and be careful. And it’s interesting that if if you’re a student of history, of which I am, and I didn’t know about this till I started researching it, that in 1900,

the previous Pope Leo I 14th wrote a similar thing about the industrial revolution. We were just moving to automobiles from trains and electricity was becoming more and more commonplace.

Um, so things were becoming automated and and this was talked about as well as becoming a a scourge on society that we

were all going to be without a job or, you know, we wouldn’t know what to do with ourselves. Well, yes, we did go from a 60hour work week down to a

40-hour work week. And with AI, we may go from a 40-hour work week down to a 30-hour work week, but standard of

living improved. The average poverty level shrank, and this is something that

should be embraced and can be used for the good of all. All right? You don’t have to replace workers with it. You can

expand and do other things with those workers with the efficiencies created.

people immediately go to, oh, everyone’s going to be without a job. I’m not saying that won’t happen. I mean, I don’t think there’ll be zero employment.

There may be a a cultural shift while we adopt this. And I’m looking 10 12 years in the future when we have robots and

nobody even drives a car anymore other than collectors. You know, there will be a a monumental shift in jobs. But, you

know, there’s no there’s no uh carriage drivers now either. There are not nearly as many horse tenders. You know, you don’t have people that shovel and

deliver coal to your house anymore. But those people found jobs, you know. So, I think the same will happen.

Okay. Another hot take from you, please.

On the flip side, what’s working right right now in our space and direct mail?

Uh obviously and everybody agrees on this is you’ve got to combine direct mail with some of these other technologies you know whether it’s it’s

printing and embellishment or sorting kind of efficiencies or it’s tracking and lift technology such as I’m in you

know that will improve the results of direct mail keep people buying it and it’s really a matter of getting the next

generation to understand the value of this type of marketing.

the digital only space is crowded. Um if you are not using some of these workflows or efficiencies including AI

in your in your plants then that needs to be looked at too you know whether it’s quality improvement or efficiencies

in movement through automation. These things need to be embraced and looked at you know yes there’ll be consolidation

and you know who will do the consolidation? the guys that are using these technologies to make their shops more efficient and make their direct mail better.

Excellent, Brad. And um I’m gonna sort of close off a little bit on what’s coming from you guys. You have a lot of exciting developments coming. You’ve

mentioned quite a few of them today. Uh version 1.1 I’m hearing um might have already been out by now. It’s out today.

Yeah. Oh, today. Wow. Good timing. Good timing.

It was yesterday. It was yesterday. uh with competitive intelligence retesting.

We’ve added parameters and we’ve got a new model training update with a another three months of the most recent direct mail campaigns that only improves the model.

We are releasing today a completely intuitive plain language search through

who’s mailing what. So you used to have to click boxes for filters. Now you just type on it like like a chat GPT. You

say, “Show me all the uh mailings from universities in the Northeast in the last 30 days.” You just type it in.

It’ll do that now.

All right. Show me the trends of of cell phone companies in the last six months.

It’ll do it. Before it was a series of filters and outputs and make your own charts. Now we we’ve AI first it. And

even our direct mail 2.0, I know all the dashboards now are AI enabled to benchmark how you’re doing compared to everyone else in your category uh what

are your competitors doing it’s all there real time through AI so like I say everything is pretty up todate and if

you haven’t looked at our platforms in a while and in fact in a week you should take another look at them. Wow.

We we we are anticipating a whole another roll out before the pavilion in September by the way.

Wow. Okay. And I guess people can see that they come and see you at the pavilion this September, Las Vegas.

Got to travel a bit this time, Brad. Um, but listen, thank you so much for coming on to the uh software matrix here on

Print Island. We’d love to have you back on to hear about your platform again in the future and and what what you’re doing for the industry. Uh, thank you so

much for sharing that insight and uh, keep up the good fight and we’ll see you in September. Definitely, Chris. It was a pleasure. Anytime.