Editor’s note: Javier Rojas, managing director of Kennet Partners, conducts discussions with founders whose capital-efficient businesses are recognized as “return leaders.” In this article he shares with SandHill readers his conversation with Joe Costello, the former CEO of SDA/Cadence Design Systems Inc. Cadence used corporate crowdsourcing, augmented with venture capital, to launch the company in 1988. It became a billion-dollar revenue business in the computer-aided design (CAD) space.
Javier Rojas: Customer funding sounds exciting but how does this work successfully? A savvy executive, (and magnificent storyteller), Joe joined the company pre revenue and then led Cadence through a growth phase shortly after it became public to almost $1 billion in revenue. He developed great insights as to what makes corporate-funded startups possible and successful. Here is his story.
Javier: How did you join Cadence?
Joe Costello: Cadence was formed by the merger of two similarly sized companies, ECAD and SDA, after which I became the CEO for the combined company. I joined SDA after it had started operations, but pre-revenue. My path there was serendipitous. I had worked for the founder, James Solomon, at National Semiconductor, but we both left and went in different directions. I had always thought the idea of a visual design platform for chip design as a crazy business idea — it was too far ahead of its time, and even if successful it was not a big market. But after spending time with James and his team, I realized that his vision for future chip design was brilliant and so I joined.
How did you fund the company?
Joe: Jim tried to start the company within National Semiconductor, where he worked. But the CEO, Charlie Spork, realized it wouldn’t be successful in house. So he offered Jim a small investment to start the business on his own condition: to seek venture funding. After deciding to work with Don Lucas, an investor and National Semiconductor board member, they agreed that they would get a majority of their first round from corporate customers.
This was very wise because it was not only going to get them money but also a first customer and the critical feedback they needed about the software to make it a viable product. In the end they raised $1.5 million each from National Semiconductor, Ericsson, Harris Corporation and GE. The raise was about $11.5 million with the majority coming from corporate investors.
The things we learned from our corporate investors ended up becoming even more valuable than the investment. It turns out that our model for working with those four investors became somewhat of a blueprint for our partnership approach to large deals.
Were all the corporate investors good partners?
Joe: Actually they all wanted to be, which is why they invested. But we found that we needed three things to make these customer investment deals work: desire, resources and a systemfor incorporating the results in the business. Harris had all three. Ericsson had the desire and the resources, but it didn’t have a way to inject the technology into its actual product flow. GE changed focus and helped us develop skills that became more valuable later. NatSemi was not great at providing input and guidance due to company culture.
So how did you take those lessons and build your business?
Joe: We took this whole model to the next level, which was the original funding idea. At the time we were burning almost $600K a month and eventually needed to raise more money. Instead of selling/raising a million bucks a month, I figured we should go big: a partnership that included a deeper access to technology where the companies could integrate their own tools.
Jim told me GE was excited about their relationship with Toshiba and he gave me that account to pitch. The head of engineering and I pitched the Toshiba guys when they came to California. We had a really good discussion and then they invited us to Japan to talk about this technology partnership idea. We arrived in Japan, and it turned out to be a three-day meeting to delve into the technology and understand it. On the third day we would have to talk about price.
We eventually realized that Toshiba really needed our technology more so than we imagined. This would be a multi-year deal where they would receive different categories of value over a period of time. When we added it all up, it was $27 million dollars. Surprisingly, they did not fall off their chairs and they didn’t throw up. We eventually signed the customer deal with Toshiba where they didn’t get a single share or any discounts, and it was pretty much at $27 million.
What convinced them were two pieces: a) our framework where we would be putting tools in and they would put their own tools where they could build a customized system for Toshiba and b) that we had a very good system for standard cell place in route. Like Jim, Toshiba had a visionary, Dr. Koyama. When we came back there were two other accounts that came after: one was SGS-Thomson and the other was Kawasaki Steel, both about the same size — over $75 million in revenue over a number of years. Success!
What do you think are the key criteria for the model to work in other markets today?
Joe: All the rage right now is crowdsourcing for funding consumer products. And in some ways what we did was corporate crowdsourcing. It comes down to a great idea and a subset of people realizing they need it. Then you have to have someone who has the resources. That combination is unusual. It is very hard to sell when there are a lot of packaged solutions because it is easier to take something off the shelf; it’s less effort and there is no risk. You have to find that severe pain point where there seems to be no solution; that is when this corporate partnership works best.
Also, you must find customers that see a clear opportunity but no bridge to it. In those days, all of these companies in the semiconductor market built their own tools. They had tried doing it in house and realized that they couldn’t do it all themselves. And it was an underserved market.
At the time that SDA was started, other EDA companies that had been funded in the Silicon Valley area were Daisy Systems, Valid Logic and Mentor Graphics. Those were the big three, and all of them had been funded and were ahead of us because they were building the complete turnkey system, which was the model. At the time, all that money was going into those three companies, not for the semiconductor designers/IC design, which was Jim’s insight. The problem was that IC was seen as a niche.
So it sounds like there were four market characteristics that made corporate crowdsourcing possible for you. The first is to focus on a small but growing market. If the market is too large, customers will likely find existing solutions.
Joe: Yes, IC design was a very small market when we started.
Next, find a problem corporate customers have today. This should be a problem they are trying to address in house because it’s necessary for their solution but not core to them. They need to know they have the problem and know they can’t solve it well.
Joe: That’s right; all of SDA/Cadence customers had gone through this learning curve so they were receptive to our help.
The third characteristic that made crowdsourcing possible for your company is that it was context, not core. The problem needs to be important to a company’s success, so they have the desire to fix it, but not core to their offering so they feel they would prefer not to own the solution.
Joe: Exactly, the chip builders needed our solution but recognized it was a different skill set than where they needed to focus.
Your experience shows that the fourth important factor making crowdsourcing possible is the necessity of bringing expertise/trust. Customers will only bet on you if you can credibly solve their problem.
Joe: This was key. Harris and the others, including me, bet on SDA/Cadence because they knew Jim was a credible visionary and they had faith that he could solve the problem.
Javier Rojas is a managing director of Kennet Partners in Silicon Valley. Kennet provides growth equity capital to bootstrapped companies in the United States and in Europe. Kennet focuses on capital-efficient businesses with annual revenues of $5M – $50M. Kennet works with founders to build high-value outcomes and preserve equity value through capital-efficient growth strategies. For more articles on entrepreneurship go to Javier Rojas’s blog at seekinggrowth.typepad.com and click here for his articles on SandHill.com.