Read the print edition

Joe Zawadzki, CEO, MediaMath

Risk aversion is a subjective concept. For Joe Zawadzki, missing opportunities by playing it safe is the biggest risk of all. His father was an immigrant entrepreneur who fled Poland with little more than a watch and a couple bucks and built a successful land development business in Canada. While some of the leaps Zawadzki has taken with his company—and his own money—may seem like significant risks to some, they pale in comparison to those of his dad’s. 

The self-described technologist was an English major and a teaching fellow in cosmology, set theory, and the history of science at Harvard University. After graduating, he worked as an investment banker. In that role, he focused on quantitative analysis, market dynamics, and incorporating new technologies into real estate investment. While in this role, he had the idea for Poindexter Systems, which later became [x+1], a marketing optimization company that he launched in 1999. While Zawadzki had professional experience with both quantitative analysis and technology, he had no background in marketing. His decision to start a marketing technology business was based on his experience as a consumer.

“The theoretical experience of being a consumer online was great and amazing and really smart, data-driven, and intelligent,” said Zawadzki. “But the actual experience was pretty bad. So that gap between theory and practice or potential energy and kinetic energy, felt like a pretty opportunistic one.” [x+1] also incubated Right Media, a digital advertising platform, which was sold to Yahoo for a reported $850 million. After leaving [x+1], Zawadzki, took about six months off and worked as an entrepreneur-in-residence for QED, a fund that ultimately invested in his next company. He also did some work for Double Click and for AOL, who had been [x+1] clients. For AOL, he helped on their supply side with their yield optimization. At Double Click he helped to create a road map for their buy-side acquisition strategy.

Insights gained from his work with those two companies helped Zawadzki solidify the model for his next business, and in the early fall of 2007, he started MediaMath. 

MediaMath is a digital media-buying platform. While Zawadzki originally used the technology to provide marketing services—a successful venture—he eventually flipped the focus to sell the platforms directly to his clients in order to realize even greater potential growth. The product has led to sizeable results for the MediaMath team. In 2011, the business earned nearly $78 million in revenue and has achieved triple-digit growth year-over-year. 

Recently, Zawadzki spoke with executive editor Daria Meoli about destroying a successful service business to create an even more successful product, running an employee-owned business, and what keeps him up at night. 

Daria Meoli: How was the launch of MediaMath received in the industry? 
Joe Zawadzki: We had this idea and it had all of the underpinnings to become something pretty transformative. Ad exchanges were getting bought by major portals and media companies. Google bought Double Click when they launched their ad exchange. Yahoo bought Right Media. Microsoft and AOL also bought exchanges. All of this happened in a sixth-month period around the spring and summer of 2007. 

There were a bunch of data exchanges that started, as well. So, we had what the founding team and I viewed as the raw ingredients for the professionalization of marketing. There were all of these technology-connected systems connecting media and data, but you had no sophisticated buyers. Everybody was basically picking up the phone and calling their rep at a publisher and saying, “I want to buy 10 million impressions this month at four dollars [cost per engagement],” It was very manual, and it was basically traditional, digital marketing applied to what could be a much more sophisticated landscape. We knew we could be the folks who connect all that stuff together and make it consumable by [media] buyers and traders. 

We had a really clear vision of this business. My first business was characterized by lots of pivots. We had a handful of key clients that were always doing really bleeding-edge things and we were learning the market from the inside. With MediaMath, I think we had a pretty strong view of how both our business and the broader market would evolve over a half decade.

DM: Do you think that was because you had a stronger concept with MediaMath from the beginning or because you personally had more experience under your belt?
JZ: I had seven years of experience in the trenches. My first business was site optimization for advertisers. So, that was all about learning what people do on your website, and then getting the right people onto the website. It was ultimately based on the idea of big data. It was audience targeting, segmentation, optimization; all those concepts that all of four people wanted to talk about in 2000. In 2013, it’s a standard part of the industry. 

That demarcation around the 2006/2007 timeframe is when we had enough experience to be able to predict how the market would unfold. Because I was so wildly off in 1999, right? I was out here saying, saying, “Optimization! It’s a great idea!” People just respond, “Yeah? No, come back to me in nine years. Part of the job as an entrepreneur is to get better and better at predicting saleability. You can have good ideas, but then you’ve got to figure out, realistically, is the market ready for them? As you start getting better at predicting, you actually can start influencing the market to be more ready for your ideas. Because you’re starting to predict the market, you are able to create these change-agent moments. 

I think we had a pretty good road map for MediaMath. Let’s start by building a technology, but then use it on the client’s behalf in order to show them that you can get transformative results. As you show them this approach to marketing, you can say, “Oh, and by the way, you can do this yourself.” Then, we licensed our clients the technology we had been using to service them. Once you create the need, then you deliver the solution. That was the road map for the business from 2007 to 2009. Then in 2009, we decided that we didn’t need to service it for people anymore. Now the MediaMath business is a platform business that we license to agencies, marketing service firms, and other intermediaries.

DM: How quick to profitability was MediaMath?
JZ: We profited quickly because of the relationships we had when we started. Our first run of financing was late 2008. We needed to raise enough capital to really professionalize the engineering [of the platform] so we could deliver something that we weren’t just using for ourselves.

We raised our first round of venture capital on the strength of a pretty good service 
business. We used it to make the appropriate investment in becoming a true product–platform business. We probably were in the red for three years as we were building up the technology and getting enough licensing that the lower-margin technology business made up for the dollars we put into it. We were profitable for all of last year and then profitable this year. 

Comfort in Constant Change

DM: Once you started to see the marketing services business grow, was switching the focus to selling the platform as a product a difficult call to make? 
JZ: It’s creative destruction. Our managed service business was growing really, really quickly, but the important part of our growth was having the courage of our convictions to say, “This looks fine now, but if we want to grow this way, we’ve got to do things that are disruptive. Things like spinning off businesses, acquiring other businesses, and factoring them. All of that leads 
to what feels like constant change to the organization. Every six months, we need to look at where we’ve reached a certain plateau, and identify how to get to the next base camp up to the summit of Everest. The things that got us to this point aren’t the things that are going to get us to the next level.

As long as you can do things that are disruptive to what feels like a good business in order to become a great business, then that’s your reinvention and it will keep your business working. That’s how you go from a good feature to a good product to a good company to a good enterprise

DM: Do you ever get resistance to this constant change from employees?
JZ: Sure. All of the time.

DM: How do you communicate changes in a way to excite instead of intimidate?
JZ: What makes for a good marketing campaign can actually be applied to organizations, as well. How do you create a test-and-learn environment? How do you say to the people in your organization, “Okay, here’s the thesis. Here’s the investment that you need to make, and here’s when the payoff comes.” Then, show people over time that there’s a method to the madness. The first time you disrupt things, you want to make a fairly short jump, with a fairly quick ROI. You want to see return quickly, because then employees will see the process work. The first two shifts feel really terrifying, but if you keep making shifts, there becomes a track record, and people will get the feeling of, “Okay, we’ve done this now nine times in our organization and the tenth one feels much less scary.” People will be able to rally around the fact that the organization needs to reorient itself and they will become more willing to do that. 

DM: Is reorienting a working business ever a problem with investors? 
JZ: I really over-indexed on picking the right set of partners for this business. I didn’t do that in my first business, and part of the reason for that was a function of the times. In my first business, there were moments where we literally had toaise money to survive, and raising money when you need to is a terrible time to raise money. When you do that, you can’t be choosy about your partners. I ended up with a business that had four different investors, all of whom had different conceptions of the business. That just led to pain in the boardroom and in our dynamics. 

So with MediaMath, I spent a lot of time figuring out which were the right firms to work with and checking references. I spent time with other entrepreneurs these firms invested in, but I also spoke to references from companies where things went bad. I actually want to know what an investor is like on the bad days. After that process, we ended up with a board and a set of investors who are in it with us. This came up again recently when we decided we needed to go international more quickly. We were asking investors to essentially double down to meet the next goal. And all of them came back unanimously and said, “Yeah, we get it. We’re in.”  

Faster Moving Markets

DM: Why was it important for you to grow internationally? 
JZ: There was a clear demand. Our business tends to index to Fortune 500, and those businesses want to make global decisions. They don’t want to make local market decisions for something as important as what operating system they are going to base their marketing on. They want to make that kind of decision once and have it federated. In some cases, the decision to ramp up internationally was reactive in the sense that clients said, “You need to be in the UK, and then you need to be in Brazil, and then you need to be in Japan.” It’s great being the first or second company of the category in a market, because incumbency plays to your favor. Once you’re in and you’re the solution people are using, it’s harder to switch. We also can leapfrog the older markets to some extent because they don’t have to go through the same learning curves. So, in many cases, international markets are bellwethers now of where the US will be, because they just move faster. 

DM: Did you experience any growing pains going overseas?
JZ: Yeah. Making sure the business feels like a global business and not like a New York business with subsidiaries or satellite offices has been really important. International markets are as important and often more innovative, so information can’t just be broadcast from New York. We have to have a flow of data back. It can’t be that stuff happens in New York and you syndicate it out to the rest of the world. You have to be receptive to good ideas that come from Australia, and use those to make a better product in New York. 

And it was simple things, like not having real-time communications and time-zone stuff that was a challenge. We ended up making a big investment to have video conferencing that’s always on and actually works. It’s the same system they have at Facebook and Google. Another thing we do is the whole team makes a lot of trips to each other’s offices, because you have to have the DNA commingling. You want people to know who’s at the other end of an email and who’s on the other end of the phone when they do call. You need to know it’s not just some random person. It’s somebody that you’ve hung out with and had a drink with. 

Creating a Change Culture

DM: How has your role evolved as the company has grown?
JZ: I think very consciously about trying not to hit my Peter Principle as an executive. The Peter Principle is reaching the level of your incompetence. In the early years, my role was very responsible for direct sales. And in the really early days, I was the first trader. I was the one that was actually logging in and managing campaigns with sometimes disastrous effects. 

I remember when the business had just got off the ground and we had our first client. I had set up a bunch of these strategies at 11 o’clock the night before going to my wife’s family’s Thanksgiving reunion in Iowa. I left the house at about 4 o’clock the next morning and arrived in Iowa around 4 o’clock in the afternoon. I had probably 65 voicemails from my team back in New York, because the strategies I had set up were going crazy. It turned out that I’d spent $70,000 on a bunch of international traffic. By the way, at this point I was self-funding, so it’s literally coming out of my pocket. It was possible that I had just lost $70,000 of my own money on that day.

I descended into the basement of my sister-in-law’s house. When I left my last business, I had left a lot of money on the table to get out, so I was in this precarious spot. I spent the next four days pulling the company out of this hole. In the end, the campaign worked spectacularly well and the client was really happy. But I had that really bad four-day period of questioning my whole life up to that point. It now has become a rite of passage. 

Now, at least 50 percent of my job is recruiting. That’s a big, important part. I’m a company spokesperson and I try to anticipate the more strategic road map as opposed to the tactical one. And usually, I like it. Sometimes I like to get back on the road and hear what’s happening in the field and how clients are thinking about things and how suppliers are thinking about things. But generally speaking, I try to figure out how to build a scalable organization. It’s thinking about the second-order problem. It’s a lot about culture and values, bringing the right people in, and making sure that processes exist to support the business without being bureaucratic in nature. It’s thinking about how you ferment entrepreneurialism so your people feel like they take ownership in the business and the outcomes.It’s a lot of meta thinking.

DM: How have you handled the transition from being so focused on product and the market to culture and recruiting?
JZ: One expression that I heard a bunch of years back that I like is, “Management is shifting bottlenecks.” You find the biggest problem and that’s where you put your attention. I’ve been super blessed in having an amazing team, and that’s not a platitude. The only thing I manage to do is recruit, hire, and marry above my station.

Management is also about getting people that are as strong as or stronger than you in their respective domain to care about your business as much as you do. I think that’s one of the transition states for business. People often struggle working in founder businesses when that founder can’t let go or can’t let people take ownership. I think that caps the growth of the business. 

We’re trying to build something of enduring enterprise value. We’ve never been a build-to-flip business. It’s never been about starting something to get it bought for $50 million. We have an opportunity to transform an industry. If you remind yourself that’s the goal and that’s what’s possible, then you start thinking, what are the things that are impediments to reaching that destination? And if one of the impediments is you, if you are too close to the business and you are not able to attract people that want ownership, then you’ve got to fix that.

DM: You said you didn’t start MediaMath to flip it, but did you have some kind of exit in mind? 
JZ: We’ve always had good secular tailwinds, and I’ve certainly said that if we start seeing headwinds as opposed to tailwinds, we’re going to re-evaluate. So, then you ask yourself, do you have that? Are you at the front of a continuing trend in an industry? Do you like your team? Do you like coming to work? Do you like the folks around you, and can your team go the distance? And if the answer to those questions is yes, you should have a bias toward continuing to grow business. And all three of those things happen to be true, and therefore, we’re going to keep growing the business.

I do think about how to facilitate liquidity stops along the way. How do you make sure people get rewarded for the growth that’s been created this year versus last year, and so on. We’re a largely employee-held business. 

DM: Why did you structure the business to be largely employee owned?
JZ: The alternative is a challenge. Investors, by definition, won’t be as close to the business as the people in it, and sometimes you end up with a misalignment as a result. It’s very normal and natural for investors to have fund life cycles. They often have a particular thesis that they want to maintain as the business evolves that’s no longer core to their investment philosophy. The more that you have that, the more challenging it is to keep everybody aligned and do the things that are right for the business as the business is today. 
It is important for us to make sure the people closest to the business feel more like owners than employees; so, everybody has equity. We have been very careful about managing the cap table to make sure that everyone is aligned in terms of incentives and so on. I think that’s led to more flexibility, more nimbleness, than most companies of our ilk

DM:  What keeps you up at night?
JZ:  I do a lot of management by clichés, actually, and one of them is, “Unrealized gains are as bad as losses.” I think in a risk-averse organization, losses feel really painful, whereas unrealized gains don’t. But I view them as equally painful. It is acutely painful for me when our kinetic energy, that is the thing we’re actually doing, isn’t matched to our potential energy, the things we could be doing. Basically, every day I worry about whether we are achieving our potential. Are we doing the things to end up where we want to be, as one of the two or three really transformative companies? And that’s what I worry about. I worry about missing that and just ending up a good company, as opposed to a great company.