Matt Blumberg, Markup AI: Restarting a legacy NLP company into an agentic AI enterprise content control layer

Welcome to another CEO wisdom podcast. We have Matt Blumberg with us today. He's CEO of Markup AI. It's Content Guardian Agent, so we're going to double-click, on that what that means nowadays. Uh he's a multi-time founder, CEO of tech companies from zero to a hundred million dollars, the author of uh Startup CEO, Startup CXO, and Startup Boards. He also hosts a podcast called The Daily Bostor.

And country over self. Uh intriguing. Matt, welcome to the podcast. Tell us a bit more about yourself and what you've been up to. >> Yeah, thanks, Charles. Uh happy to be here. Uh yeah, so I look, I've been running tech businesses going back to 1995, sort of the beginning of the internet. Um businesses in lots of different spaces, uh the movie business, the email business, uh the the recruiting business, and now the um sort of content production AI business.

Uh so uh lots of different experiences, public, private, uh venture-backed, private equity owned. Um you know, kind of full range of things. And then, as you mentioned, some good extracurriculars. Uh I've written a few books, uh sit on some boards, I've done a couple podcasts. Um Some business-focused, and then some uh American history-focused stuff, too.

>> Interesting. Let's start with uh your education. You have Princeton, but urban planning? Like, what graduated from urban planning and ended up being a tech CEO, a very successful one. So, tell me a bit more about uh that early life of yours. >> Yeah, you know, um my parents always said, like, doesn't matter what you study in college, study something you want to study.

>> As long as it's Princeton. >> [laughter] >> Uh study what you want to study, learn how to be a a critical thinker, uh learn how to learn, uh learn people, and you can do anything you want with it. And um urban planning for me was a was a mixture of a couple different disciplines that I really liked. Uh one was uh civil engineering, a kind of architecture, so the building, which I guess translates into entrepreneurship.

Uh the other was public policy, so that major combined them and um you know, interdisciplinary. My whole life it feels like it's been interdisciplinary, so I guess that started in college. >> And what was your first company? What was your first stint in the the startup and tech world? >> Yeah, I joined an existing company called MovieFone uh 777 Films uh in 1995.

The company had been around for 6 years, so I didn't found it. It was an interactive telephone business and it was pre-internet. Uh and I was hired to figure out what the internet was and ended up being the uh kind of product manager and then general manager of moviefone.com, which uh in 1995 at some point was like a top 50 web property. Um first place you could buy a movie ticket online, first place you could uh get uh customized movie listings and e-commerce through email.

Uh lots of firsts in the very early days of the internet. A very, very disruptive business to old media, to newspapers and television uh at the time. >> And you exited just on time for half a billion uh right before the the bust. >> Just on time, yeah. And I again, I not it wasn't wasn't my business. I was a happy member of the executive team, uh but it was great experience and kind of a good uh sort of first uh entrepreneurship experience of building a business inside of another company before I started my first company.

>> And why didn't the CEO {slash} founder didn't give you any shares back then? >> no. I mean I I I had options just like other employees and and did very well for a 20-something-year-old at the time, but uh it was a big headline number and it was not my headline number. >> Yeah, yeah. How did that change your mentality? You were like, oh that guy just exited and he's probably worth tens of millions now.

How did that influence your mindset? >> You know, it was it was never about the exit and the money. Um quite frankly, the thing that uh uh I really took away from that experience was wanting to start my own business um more because I had uh a lot of interest in building a different kind of workplace and a different kind of culture um than anything else.

And that's really what influenced me to go do my own thing. That and uh my you know, sort of my entrepreneurial role model in life is my dad who was also uh a founder. Um, really had me give gave me the bug to go do something on my own more than it was like, you know, starry-eyed IPO exit thinking. >> Now, why did it take you so long to launch your your first stuff?

Because yeah, I was my dad was an entrepreneur, too, and off the gates I've started something. Um, seems like you you followed more like the traditional path, right? Like getting high-paid job. Uh did you did your dad push security on you or did he sort of push you to also take risks? >> No, no, he he he didn't he didn't actually never had an agenda with me.

Um, it was uh uh quite frankly, starting a business at 29 which is when I started Return Path in 1999 was still considered early at the time. Now it's not. Um, but uh you know, that was before everything. >> [laughter] >> I mean, it wasn't quite before the internet, but it was let's put it this way, it was before the cloud. Like we started a tech company, the first thing you did was get a data center and buy servers.

And um >> Physical servers. >> You know, you had to you raise more money to get started. And so the the the barriers were higher. >> Yes. >> And uh yeah, I had had a a few jobs, but I wouldn't describe any of them as I wouldn't describe what I had as a long career in high-paying jobs. >> Right, yeah. Um, maybe I read a couple ones on your LinkedIn as associate and consultant for Mercer.

But your first official one was Return Path, right? >> That was the first company I started, correct. >> Yes. Yeah, the I'm suc- quite successful. Tell us a bit more about that one. >> Yeah, that was a great journey. Um I started the business in 1999, sold it in 2019. So, it was a 20-year run. Um we uh were uh kind of pioneers in a couple of uh of corners of the email uh infrastructure sector.

Um we built some really unique data and analytics uh assets. Uh and we had an enterprise global enterprise SaaS business at the end that had about 4,000 customers. Uh we had about 500 employees. We were in 12 places around the world. And um uh really well-known brand and a really, really great place to work. Uh which was part of what I wanted to do from the beginning.

So, um you know, we won a lot a lot of awards uh for our culture and being a great place to work. And that part of the success was just as important to me as building a building a large business and a scaled business. >> Tell us more about the culture. Seem pretty fixated on it for a reason, right? Because that's that was like >> Yes, it's uh it's been instrumental to your success so far.

So, double click on that and like systems that you've built to uh attract very aligned humans in your company. >> Yeah, the um the philosophy we had at Return Path and and the way it was articulated there may be a little different than I've articulated it in subsequent companies, but it's the same uh same general just of things. Um is what I call people first.

It's very simple. Um you have a lot of stakeholders in a business. Uh there are a lot of companies that are shareholder first. There are a lot of companies that are customer first. Um and uh the business was people first um with the theory that if you put people at the center, if you um hired the best people, if you developed them and grew their careers and gave them a just a great environment, a place they wanted to come in every day and connected them to the mission and connected their work to meaning, um that you would uh by definition do the best job for customers.

You would build the best products. You would have the best customer service. And then if you got this if you got that right for stakeholder number two, um it would work out for shareholders. Uh so I think that philosophy has been really important to me in my career as a leader. >> And you lasted 20 years in return path. I'm you I'm saying lasted cuz that's an exceptional one.

20 years. Uh what did you survived? Uh what challenges came to you and yeah, like how much near deaths in the in that 20 year span? >> I mean, let's start with uh let's start with 9/11, the dot-com meltdown, the global financial crisis, take your pick. Um but lots of, you know, lots of near death experiences. It was not a linear journey of 20 years, you know, smooth and up into the right.

And there were yeah, there were a number of times where I feel like we looked death in the face. And quite frankly, I'll bring the point back to culture a little bit, which is I think there were probably a bunch of points where the business might have died except that the company was really strong. >> Yeah, that's fascinating, right? Like employees are still there like holding each other up.

Um I mean, the counter argument is like do they have other options? Um but but uh >> We had really good um retention and tenure of really strong employees because people wanted to be there and they wanted the company to be successful and they they could get on board with a pivot. Pivots, you know, were not scary, you know. >> But what did you do when like there was like almost no cash in the coffers and like like not going to meet payroll?

I guess you're a nice guy, nice boss to have, I guess, and people built uh shared values and have fun working there. So, it's like, "Oh, let's just wait. Uh we trust Matt. He's going to figure it out." And 2 3 weeks later it was figured out. Or what were these moments like? >> Probably the right way to describe it. I mean, the the moments were um you know, uh harder than you know, behind the scenes than uh you know, than than maybe were on my face every day.

But um you know, cultures like that have a lot of shared trust. The same way like I'm not an engineer, and I trust that the engineers are going to do a good job. Um you know, I trust that they're making smart architectural decisions and writing efficient code and making good decisions about technical debt. You know, I could say the same thing about almost any role in the company, and I think people felt that way about me.

Like, you know, that's one of the things Matt handles, and it'll work. >> Did you give them equity? >> Uh every every single employee. >> There you go. Yeah, I I love that. I I wish like all the >> That's That's been true of That's been true of every company of mine. Um including including the current one. >> No one was doing that back in your days.

Like, very few people gave equity to uh pretty much everyone. >> It was less normal then, for sure. >> Yeah. And it's complicated also to give them equity, right? Like, financially you need to pay for the lawyer papers and all of that and the complexity, the compliance. Um how did that one ended up? How did the run ended? >> I mean, it was great. We uh sold the business in 2019 to a company called Validity, uh which was a private equity-backed roll-up that in the marketing data space.

Um they bought Return Path basically to be their anchor platform. They've done a tremendous number of bolt-on acquisitions since then. I think they've done a great job of continuing the journey. Uh we handed them the baton, and they've uh continued to scale it uh really nicely. >> Uh Foundry invested in you. So, Seth Levine wouldn't have uh guy that I had on my podcast many times.

Good guy. >> Yeah. Seth Seth was not never on my board, but he and I actually were on another board together. Uh Brad Feld was the partner from Foundry, Seth partner, Fred Wilson was my one of my anchor tenant investors from Union Square Ventures, and Greg Sands from Cota Nova Venture Capital. Those three were at Return Path. They were also on my cap table at Bolster the business afterwards along with Scott Dorsey from High Alpha.

And yeah, great great group of venture investors. >> And SoftBank? Did you speak to SoftBank? >> no, no, not not quite. It was SoftBank Venture Capital which was Brad Feld which became Foundry basically. >> Oh, okay, got it. That's SoftBank. What [snorts] happened after? That brings us to 2020 which is like not so long ago. What Bolster I guess, right?

>> Yeah, 2020 a group of us left Return Path after the acquisition and we started another company together called Bolster. The premise of Bolster was helping other founders scale themselves, scale their teams, and scale their boards. So the idea was to basically create a different kind of new kind of search firm that was designed really to be a platform not a high services high touch search firm.

And we did that for about four years. There were a lot of aspects of it that worked really really well. We placed about 500 people at startups all across the country and even some internationally. About 100 of those were independent board members. We did a lot of great intellectual property development, but we had a really hard time building it into a scalable platform a scalable technology business.

The gravitational pull was much more to to being a services business like everybody else in in the search space. And we just didn't we didn't we weren't into that. That wasn't what our DNA was. That was a lot harder to do. There are a lot of entrenched relationships in search. So we decided to wind that business down in 2020 2024 it just wasn't going to produce the the kind of return that our venture investors wanted or that we that we wanted, so we returned some capital to shareholders and there's still some remnants of it out there.

There are a couple of my co-founders that do some of that some of that work freelance, so you still see the brand and the website, but but the business doesn't really operate anymore. >> But now you're all in full-time in Mark Up, right? Like why why did you make that decision? >> I am all in full-time on Mark Up AI. This was a really interesting opportunity.

This is a business that I didn't start, but it's a business that I'm restarting. So I was recruited in and was able to bring a management team in with me, a number of people that I've worked with at my prior companies. And we were brought in to to basically pivot a company that had a legacy software product in natural language processing and a lot of business from Fortune 500 customers that had been with them for 10, 15, 20 years.

And pivot that into an agentic AI business. So on the one hand we're building a native AI startup, which is a lot of fun and puts us kind of right in the in the middle of the most exciting part of tech right now. But we have the the sort of the strength of a legacy business behind us. We're about to launch our platform in less than 2 weeks and first order of business is to migrate our clients from our old stack to our new stack and then uh co-equal first order of business is to get new clients in the door.

We've already got about 200 customers on the early access list and we're really excited about it. The premise behind the business is basically you know, job one for large language models was code. And there's you know, a well-known and rich ecosystem of vendors and platforms that serve the coding assistant world. We think job two is going to be content.

And you know, companies are increasingly turning to LLMs to produce enterprise content. Um and what that means is that they're getting more content at lower quality um in an environment of of changing stand- rapidly changing standards. So, anyone I talk to that touches content in an enterprise says, "Oh, yeah, my job is now editor because anything I get out of a model isn't quite right." And the Agentyc platform that we're bringing to market is basically a a control layer for content.

So, uh it's agents that help companies keep their content on brand, compliant with laws and regulations, accurate and fact-checked, and uh ready to go for AI search and AI discovery. >> So, who recruited you in the CEO role and were you dropped on the battlefield? Or were you maybe still from an helicopter? >> I got headhunted by a headhunter. Uh the the same way that we had kind of been doing it at Bolster >> You got captured.

Why Why did they pick you? Like, I mean, you have a uh very strong track record, right? Like, why were you the right fit for that company? >> I think the board um knew that they needed someone that was not a not only able to scale an existing business cuz the business wasn't scaling anymore. The business was in decline. Um and I think they preferred someone who was a business builder and had seen zero to 50 as much as they had seen 50 to 100.

Um not necessarily a turnaround person, although the business is a little bit of a turnaround cuz it's not a it's not a stable mature space, which is where you think of turnaround. This really needed to be a restart. Um the company was also um originally based in Germany and they private equity owners who are German um wanted a a US-based CEO. Company had some US employees, had most of its clients in the US, but really wasn't part of the US tech ecosystem.

Um and so, part of pivoting the business was also pivoting it to uh to the US and to kind of the the startup and growth environment. >> It's pretty interesting though, like all the culture at the right place, right time because you were >> right time. Yeah, I mean the the real attraction for me was was less doing a turnaround, although that's been really interesting.

It was more the opportunity to work in the AI space, but to do it from a position of having um you know, 100 to 150 um you know, global 2000 enterprise customers. So, um yeah, you can start an AI business from scratch and you can do well with, you know, getting small early adopters, but I think what we're going to have here at the end of the the turnaround journey is an agentic native agentic AI business with Fortune 500 clients on it.

And uh that's a great foundation for us to build on. >> So, what's your GTM playbook to get these guys to try you and use you? >> Well, the existing clients, uh the playbook is less GTM and more about product and functionality. Basically, we replicated the functionality of the old system in the new system as a as a series of agents. Um and uh the migration path for them is going to be very straightforward.

Um because it's the same in some respects the same system that they're used to using, but it's uh housed in a in a much better um much more contemporary body. Um So, that's a little different than the go-to-market playbook for new, which is going to look much more like you would expect um a new uh native AI company to do. Free trial, you know, start with early adopters and disruptors, start with an atomic unit of one or a small team, land and expand, forward deploy engineers, and the like.

>> So, when you say agents, the agents would go and correct the uncorrect content of my brand? >> Correct. Yeah, the agents can be deployed uh through a few different surfaces. Um and there's a little bit of configuration up front, although that only takes a few minutes. You have to inform the agents like, what is your brand, you know, what's your brand guidelines?

Uh you know, you can upload documents that do that or point us to web pages. Um do you have a terminology dictionary or glossary? What laws do you want us to check in terms of compliance for accuracy and fact-checking? What's the source of truth? What content repositories do you want to tag for which issues as a source of truth? So there's a little bit of set up but that that is trivial.

It takes minutes. And then the way you interact with the agents we we're very surface agnostic. So quite frankly MCP server and APIs are the way that a lot of people are going to interact with the agents because we'll plug in as a tool called during the moment of content creation or do batch checking of whole repositories. But we also integrate with with content applications through a Chrome extension.

A couple of times we have a bespoke integration to some systems that are a little quirky or are still desktop based but for the most part the Chrome extension is going to do the work so that an author can use us at any point in the life cycle, content ideation, content creation and editing, batch checking. >> And for a guy like you that loves building teams from the ground up was it a trauma slash a shock to be dropped in a battlefield and probably have to let go of some people and restructure some stuff?

>> Yeah, there were some really difficult moments I dropped onto the battlefield is not a bad way of describing it at all. I will say you know this is about this is about founders so I there are a couple of lessons that I kind of take away that I think are interesting for founders. One is I think every founder should have to do a turnaround because it makes you sharper.

You do not have the luxury of dithering that you do sometimes you know as a startup. But one of the other things that's been interesting for me just for kind of mentally or psychologically is when it's your own company when you're the founder everything's personal. And when you're dropped in to fix something nothing is personal. And that's made a huge difference.

Now that doesn't mean it's hard to to say goodbye to employees, even the ones you don't know. Um you all right, you understand emotionally what you're doing to someone, but you also understand that if you don't do it, the company is going to die. Um so it's definitely challenging, a whole new new set of challenges. I think if I ever am a founder founder again, which who knows, um I think I'll be a much better one uh for having done this work.

Um but yeah, I mean it's tough to be a builder, and the first thing you have to do is is throw throw the gear into reverse. And um you know, we've taken a lot of lessons out of the playbook that our team has developed over the years for scaling up and literally done the opposite. Now, fortunately the business is stable, um and we're cash flow positive.

We're about to launch our new product, and uh hopefully we'll put the gear back into into drive and accelerate forward. >> The first I mean, I've been I've never been dropped in a company like that that that not not something I would be interesting to me as well, but I had various fires and wars I've fought in my own businesses for sure, and uh some instances would I which I had to sort of take teams, uh and the first thing I would do is find lieutenants, right?

Like find lieutenants that can be trusted and can be like directly under me, and start to execute executing stuff. What did you have to say about this? >> Yeah, for sure. I mean, I think one of the one of the things we've done well here is we've blended new team with old team. So I did bring some people in with me because there were some open roles and some some part of the nature of turnaround and moving to the US was we had to move some stuff around.

Um but there are some critical people in the organization, probably about half and half, um that have been here for a long time, uh that have the value of institutional knowledge, of client uh and um you know, who uh you know, have been essential to the success of the of the business and of the turnaround. Um and I think I think you know, the team has done a nice job of coming together around shared goals and shared values to move forward.

And it does there's never been a moment where it feels like, oh, it's the new people and the old people. >> How do you manage culture clash? Like the German culture, American culture, quite different, right? >> There were so many different points of culture clash, front of the house, back of the house, German, US, although we've got people in other countries, too.

Um old people, new people. At some point you >> Or it's Mustangs. >> Yeah, exactly. At some point you have so many different vectors that it's not you're not focused on one and you just have to get everybody focused on the task at hand. Um as I always like to say, the business is the boss. Um so, we spend time getting to know each other and getting to trust each other as humans, but everyone's focused on pulling in the same direction and we kind of we try to just set all of that stuff aside.

It doesn't matter. Doesn't matter if you're German or US. Doesn't matter if you're an engineer or in sales. Um everybody's got to pull together to make this thing a success. >> Yeah, you got to be an M badge or an AMG badge to be recruited. And that's how I'd put it. Um what's next for you? What this year, what's your focus? >> Uh launch is June 23rd as we're recording this on June 9th.

Uh so, all hands on deck to get this product launched and get adoption. And nothing else matters until the end of the year. We got to get adoption by legacy clients and adoption by new clients. We want to finish the year having moved or planned the move of every legacy client. That's about 120 large enterprises. Um and we want to get another, you know, equal number of new clients in the door.

Um you know, prove the thesis and uh you know, and start the journey forward instead of the journey backward. >> Matt, where can uh people find out more about you? >> Uh lots of places. Um my blog is startupceo.com. Um it's not really about Markup AI. It's about management leadership. I've been posting a lot about AI recently, but not so much about this company.

Um my email address is matt@markup.ai for anyone that's interested in what we're doing. Uh we're still looking for early adopters. Um product is free at the moment and uh excited to have people try it and give us feedback. >> I was Matt and Jeff.