Plano-based iTKO is a successful enterprise-software company that offers a pioneering suite of products called LISA. These products help companies “mock up,” or test, solutions for developing applications in composite and cloud environments, at enterprise speed and enterprise scale. Founded in 1999, iTKO (pronounced eye-T-K-O, short for Interactive TKO Inc.) has grown over the years to win customers including Fortune 500 banks and telecom firms as well as companies like American Airlines and JC Penney. Recently, at a Dallas forum presented by the TeXchange group, we (and the audience members) had a chance to talk with John Michelsen, 42, iTKO’s co-founder and chief technology officer, and Shridhar Mittal, 45, the firm’s president and CEO. We wanted to learn about the company’s origins, its innovative growth strategies, and its acquisition for $330 million last year by Islandia, N.Y.-based CA Technologies, a provider of IT management software.
Q: John, how did you happen to invent LISA back in the late 1990s?
John Michelsen: One of the first companies I was responsible for starting—a boutique software company—did a lot of system-integration work. We had a lot of fun building applications for customers. Every company that does that ends up doing skunk-works-type things to make their jobs work better. They do things that the industry isn’t already doing.
We started building what’s now called LISA, really in the garage, just to help us get things done, assuming that the industry leaders were going to eventually do something like this. This whole distribute/computing/message-based architecture—or service-oriented architecture—started happening. But the tooling didn’t support it; the processes were changing. We needed something, we built it, and we figured that the industry would catch up. When, by 1999, it still had not, we decided we would actually start this company and get it going.
What is LISA, exactly?
JM: It’s software. Client companies spend an enormous amount of time and labor building, testing, and releasing their own custom software, or heavily integrating and customizing packaged software. The risk of failure in production is very severe. And, these days, it’s a brand issue. When the company website doesn’t work, or when your iPhone app doesn’t work, customers leave. So we produce software that helps solve problems in that type of environment.
Did you have a eureka moment with LISA, when you said, “I’ve discovered something here”?
JM: Well, no, but necessity is the mother of invention. So we had the opposite of a eureka moment. We had an “Oh, ****!” moment where, in about the Christmas season of 1997, we had built British Airways’ website for them. But, we didn’t have a whole lot of time to build much infrastructure for it. We built it in three months. Back in those days, a web year was three months long, right? You ran roughshod, you did everthing as fast as you possibly could, to get it out there in front of customers.
Well, I was on my way to Colorado to ski with the family, and I heard a spot on the radio for British Airways’s worldwide fare sale. That [promotion] came crashing down on our website, and it couldn’t handle the load. So, of course you go to the shelf and look for the tools that can help you solve those kinds of problems—while I’m driving back from Colorado. And, there just wasn’t anything there. So we started building what is now LISA.
Where does the name come from? Is Lisa your wife or your girlfriend?
JM: No, it’s nothing that clever. Every one of these skunk-works projects had a name that was an acronym—like JESSICA, SARA, SANDRA, LISA.
LISA was a so-called “disruptive technology.” What does that mean in terms of your industry, and how did its uniqueness help you grow over the years?
JM: Essentially, LISA helps companies save money, to get to market faster, to have higher quality in so doing. There’s a saying: ‘You want it cheap, fast, or of high quality; but you can’t have all three.’ Well, we actually do achieve all three. Our software helps customers do it faster, cheaper, and better, all at the same time. So that’s pretty useful for them. Even though we were small in the early going—we had very little in terms of marketing—we were solving a very real problem. So there was enough of a following early on to gain traction.
What were your first marketing efforts like?
JM: We actually announced the product, and went to market, at a web-services trade show in Boston the very day the U.S. invaded Iraq. We wanted mainline developers to know about our new product. Along with that, we did about six months’ worth of the goofiest, scariest-looking full-page ads you’ve ever seen, which we still keep up around the office! That cost 50 grand and got us a couple of nice articles published and some awards. We had freely available software online, and we did that up until Shridhar showed up. It was a ‘Try it; if you like it, you can buy it eventually’ offer. That was how the first customers came.
There’s a funny email that I’ll never delete, from our very first customer. It said, ‘You reply faster to the support emails than you do to the sales emails. So, do I send you the money for support, or do I have to wait for someone to ask me to buy this thing?’ Because, early on, it really wasn’t about making money. It was about getting this thing right and making sure that we actually had something that worked. We were solving a very big, very complex problem, and it took a lot of software to do it. You couldn’t go to market with something small and try to work your way up. We had to come to market with something big to start with.
How does LISA help companies save money?
JM: Today, with all of LISA’s capabilities, a customer will take 30 percent or more off their timeline to develop, test, and release a piece of software. We’ll take significant amounts of infrastructure costs down; it’s anywhere between 10 percent and 80 percent. And, at times, customers literally can show us the $30 million budgeted item they didn’t have to spend because of the use of our software. Their number of defects in production goes way, way down. In fact, one of Shridhar’s enormous contributions was helping us get customers to visualize what [those savings and improvements] would look like in terms of their business value, in advance of the purchase.
And there’s nothing else like this out there?
JM: At that time—and even today—we have a pretty open field.
How did you get your funding in the early years?
JM: It came out of thin air, actually. I’d had a successful consulting practice going, so we bootstrap-founded the company in 1999. Then I brought Ruston Vickers in as a co-founder in 2001. I built most of the product, and probably about half of the products’ code base, between ’99 and 2003-‘04. Then I spent a few years selling—not making much money, but selling— and then in ’05 decided we needed to scale up and go get outside money.
You also decided then to bring Shridhar on board. How did that decision come about?
JM: My business card reads “Chief Geek,” literally. I am still a technologist. I did not want to build a sales and marketing organization. I had started a previous company trying not to do that; when it turned out that I needed one anyway, I sold the company. So when it was time [for iTKO] to get big, I knew there was no way I was going to do this; I needed someone who had been part of a large organization, who had grown a large organization. An associate said, ‘You know, you’re going to want to replace yourself as CEO about a year or two from now.’ And I said, ‘If I’m going to do that in a year or two, why not do it now?’ He said, ‘Okay, then I’ll help you find the right guy,’ and that’s when I started looking. And we wound up selecting Shridhar.
And you were at i2 Technologies, right, Shridhar?
Shridhar Mittal: I was at i2. Well, I had decided to leave i2, and I had several opportunities I was considering.
Were you aware of John’s company at that point?
SM: I was not. I had no clue. I’d never met John before. In fact, I didn’t know much about the space at the time. When we did meet I told him that I’m a software guy, and so I know how to build software companies, etc. But I didn’t really have any idea about the particular space he was in. It was only after we met and I started understanding what John had done that I started truly understanding the company’s potential.
How big was the company in 2005 compared to where you are today?
SM: As I recall in ’05 it was just John and Ruston.And we had a part-time sales guy who was working on a commission-only basis; his previous job was selling insurance. Suffice to say he wasn’t a real enterprise-software sales guy. He was a very good sales guy, but he just wasn’t an enterprise salesman. Today, we have about 130 people.
So, what were your first moves after joining the company?
SM: When John and I spoke, the second [LISA iteration] was ready to go, and customers were already using it. So it was just a matter of positioning it correctly and selling it. I think the very first thing, before we even thought about hiring a sales rep, was deciding what market we were going after. From that, everything else followed. So, we did a full strategy session. We started thinking, ‘Okay, so we need to go after the enterprise space, and not after the mid-market.’ I’m a strong believer that you need to take it one step at a time, because eventually, we figured, this technology can be used by every single software-development company in the world—whether it has only 10 developers, or it has 10,000 developers. But we knew that we were not going to go in and attack 10,000 companies all at the same time.
After deciding on the enterprise space, it was, ‘Okay, then how do we position the product?’ Because it’s a very strong technology, and you can position it possibly 100 different ways. But the thing is to position it in a place where somebody is actually compelled to buy it—where there’s a value proposition associated with the product. Prior to 2005, the software was being sold a lot as a Java testing tool, which was its strength. The people who were doing a lot of Java development, like John and Ruston, didn’t really have the tools. So it was very popular. The only problem was that individual developers or even teams of developers would buy it possibly for $50,000, $60,000, $80,000. We wanted to go up to a market that could pay us hundreds of thousands of dollars. So we applied into a broader market called SOA [service-oriented architecture], which was an up-and-coming thing in those years. The two dominant players in software testing were IBM and HP. You can imagine trying to compete against IBM and HP. We could have taken the route of hitting them hard and saying, ‘We do the same thing that HP and IBM do, but we do [it better and cheaper.]’ But we didn’t take that route. We went down a different path, [focusing on SOA testing,] and said, ‘We are going to help you do something that HP and IBM cannot help you do.’ That, I think, was key for us. Once we got that message down, once we knew what market we were going after, we started hiring the sales force.
So up until 2005 your annual sales volume was sort of minimal?
JM: Over those two or three years it was $400,000 or so.
But once Shridhar joined you, things took off?
JM: Yeah, and it’s a funny story. As Shridhar was joining and learning, he started negotiating. Sun Microsystems was our first legit, big deal, for $200,000. And that was right on the cusp; we had to transition. Again, we had not been selling this through the support emails or sales emails. I actually refused initially when Sun said, ‘If we are going to spend hundreds of thousands of dollars, come out and visit and help us understand this.’ I said, ‘No.’ I said, ‘If you can’t figure it out from the website, from all our materials and our emails, I’m not going to sell it to you.’ So, Shridhar talked me into going once to Sun. I flew to Denver, I spent maybe three hours with them, flew home and said, ‘Okay, that’s it. It’s either buy it or forget it.’ They ended up buying, and that was our first big sale. And then we had another three or four sales in the next few months after that.
SM: That Sun deal was what validated the viability of the company to the VCs at the time. That’s how we raised our Series A in November of 2005.
How much was the Series A?
SM: The Series A was $5 million, or roughly 50 percent of the company. We got a $9.2 million valuation. What we were trying to do was basically build a sales force. We wanted to hire a couple of developers, obviously, but our real intent was to hire a good sales force, to do some marketing, to get some leads going.
At this point, where were you in terms of product development?
JM: Every generation of the product introduces a major new piece of functionality. We’ve always spent a lot of time building, and continuing to build, the product. Even though from the beginning we’ve enjoyed a distance between us and any competition, we’ve never thought that we should slow down, dial down investment, stop innovating. We have always driven even harder and expected even more. In fact, it’s always been a joke around the office that in January of every year at the development meeting, I say, ‘What we’ve done is great. We just have to work that much harder. This is the next big release.’ So, we were actually doing quite a bit with the LISA product.
And then you did successive rounds of venture capital?
SM: Yes, we got one more round of $7.5 million. And then an add-on of $5.75 million.
John, you must have overcome your reluctance to actively sell pretty quickly, then?
JM: Sort of. We spent a lot of time looking for a selling CEO. Alright, we’re good; we’ve got ourselves a great selling CEO. I say, ‘All right, Shridhar, that’s your problem; the product itself is mine. So, I’ll spend next week in Australia [talking with] customers [about the product]; I’ll be in Hong Kong, Singapore, three weeks [talking with] customers.’
That’s because you have such keen knowledge of the product, right?
JM: I think I understand the space, so I spend a lot of time talking to people who actually understand their particular problem. Our software is targeted directly at [those] software problems.
SM: I don’t think that we have a better salesman than John. The passion that comes out, the depth of knowledge, the breadth of knowledge, being able to position at the CIO level as well as downward.
Did you face many pitfalls transitioning to a much larger operation?
SM: Positioning the product was a big one, but I think we did a pretty good job at that, and then we stayed consistent with it. The real challenge was in hiring the right people, because even after we got the right people in, we made our share of professional mistakes. But the good thing we did was that when we recognized that we made a mistake, we fixed the problem.
Dallas-Fort Worth is known for the quality of its work force. Were the employees just not well-suited for your particular company?
SM: Well, it’s a small company. And people don’t always want to leave their steady jobs and go for a small company, because, what if it doesn’t work? Then they try to get their [old] job back. So, that I think was our biggest challenge.
How did you wind up attracting good workers?
SM: We like to get internal references to hire people. So we got a few, and then they got their friends, and then they got their friends, and that’s how we did it.
What is your approach in terms of developing talent?
JM: We don’t have a philosophy of 1A players, 2B players, 3C players. They’re A’s or they’re out. You can’t outrun a competitor that has a $150 million R&D budget with a team that looks just like theirs. You just can’t do it. I happen to have been exposed to a few great companies. At one of them, I ran R&D for Trilogy Software out of Austin for a couple of years. So the core of my development team is actually still sitting in Austin—all in their own home offices. There’s also a nice nucleus of the development group here in Dallas, which comes from another place I spent time, which is Eagle River/Agency.com.