A Long Conversation With John AlpineOthers Sunday, August 8th, 2010
While attending the Spatial European Forum in Frankfurt, Germany, I had the opportunity to sit down with John Alpine, Spatial’s VP of R&D, and get to know him better. I first met John at COFES 2010 in April this year. One of the outcomes of this conversation was the real story of how Spatial came about with its CAD on the Cloud technology (see “Spatial’s CAD On The Cloud – The Real Story“). John talked about his education (which I’m sure you will find a bit weird), his jobs in different companies and what went behind the scenes at CoCreate and Proficiency while trying to sell the companies. We spoke for a good 45 minutes. Here is part of the conversation.
Deelip: So what’s your story?
John: Well, actually I didn’t start out in Engineering. I started out in Anthropology. In fact, Cultural Anthropology.
Deelip: What the hell is that?
John: (Laughs) It’s a a study of human cultures. I picked the Berbers of Morrocco, a native tribe that nobody knew much about.
Deelip: Hold it right there. What is your position at Spatial again?
John: (Laughs) Well, I’m VP of R&D at Spatial.
Deelip: I get the feeling that this is going to be very interesting. Please carry on.
John: (Laughs) So I went to UCLA for my undergraduate and then went to the University of Chicago where I did my graduate work in Anthropology. Then I couldn’t get a job because it was probably a bad time for anybody in that field (laughs). So then oddly enough my father who was an aeronautics engineer for Lockheed Corporation found me a job in the company’s drafting department. But the irony is that I have probably the worst handwriting in the world. And in those days you needed to do a lot off handwritten stuff. But luckily for me, at that time, they started looking at computer aided drafting.
Deelip: When was this?
John: Oh, it was so long ago. Probably in the seventies or something like that. They were doing electrical engineering work and I worked in a group that was primarily doing microwave radar stuff. The nice things about those corporations back then was that if you wanted to get an education in a particular field they would pay for you to go back to school. Also at UCLA they allowed you to do most of the classes by television at your work site. So that meant I could take a break from work, attend a class and then go back to work. I had to visit the campus only to do my tests. So after doing my Electrical Engineering I got a real job at Lockheed and then worked for Hughes Aircraft.
Deelip: You mean Howard Hughes? I remember the movie “The Aviator”?
John: Yes, that’s the guy. In those days they were designing missile systems for the Department of Defence. I worked in a group that put the radar pieces together. One of my colleagues there used to love to work with PC’s, which were quite primitive at that time. He got tired of using mainframes for simulation of radar and microwave circuits. So he created a tiny program that ran on a PC which did some of the calculations for driving what we call S parameters. He decided to go to this Microwave and Design show his little PC software. It wasn’t really finished and crashed constantly. But he sold 200 copies in one day. So that showed the pent up demand of that kind of software. So he founded a company called Eesof which I joined and we started selling these simulators. That company really took off. So much so that we ended up competing pretty heavily with HP and eventually HP just went ahead and bought us. So that’s how I ended up in HP. I loved working there. While I was at HP I met Tillman Shawn who was a General Manager. At HP he had developed the ME10 and ME30 CAD systems. They also had something called Work Manager which was kind of a PLM system. He decided to try taking these things outside HP because this stuff wasn’t really a part of HP’s core. He actually managed to convince HP to divest. So he went independent and got some outside funding, started CoCreate and I joined him. Tillman had a vision of collaborative CAD which was probably far ahead of its time. Basically different people working on the same model from different places at the same time.
Deelip: You mean something like what people say the Cloud could be used for.
John: Yes. That was called OneSpace. Even though we got that product to work and we could see customers really get excited about it, they wouldn’t buy it. We tried a bunch of different technologies and new ways of doing things. But I finally came to the conclusion that engineers are little like designers or artists. I mean, you can’t really ask a bunch of artists to go and paint on the same canvas. It wasn’t the natural way to divide work and that concept didn’t really catch on in the market. So then we went back to the core CAD stuff and began to lay more emphasis on Direct Modeling.
Deelip: So do you think things have changed now?
John: No, I don’t think that particular fundamental has changed a lot. I don’t think designers want to work that intimately. They need to coordinate with each other at certain points but they want to work independently. So anyways, we concentrated more on Direct Modeling. But probably the main problem was the user interface. I mean, if you think of Direct Modeling you probably think its something that appeals to more naive designers because you do not have to think parametric ahead in time. And if you have a real complex user interface, it just going to be a barrier and I think they probably never got the user interface right for a long time.
Deelip: They meaning you. You were the CTO, right?
John: Yes. You know it was one of those cases where we were too customer compelled as opposed to being customer responsive. If you are too customer compelled you just keep on adding stuff that every customer requests and the software gets fatter and fatter and more and more complex. You know, there’s twenty ways to do the same thing because twenty customers asked for the same thing in different ways. I just think all CAD is going that way. People have not effectively resisted the temptation to add more and more stuff. Anyways, towards the end of my time the investors wanted to sell the company. So they brought in superb salesman called Bill Gascoigne. Bill had previously sold out SDRC.
Deelip: But what about Tillman?
John: Tillman had already retired. So we tried to shopping CoCreate around. We went to Autodesk, PTC. I mean, we went to everybody. But nobody was interested in us. So the investors told Bill to back off, build up the margins of the company, get it to be very profitable. And he did it quite elegantly. He just didn’t spend money where he thought it should not be spent and he eventually got it margins way up and then sold it to PTC for I think 250 million. And earlier they wouldn’t buy us for even 70 million.
Deelip: Wait a minute. If I understand this correctly Bill didn’t really change much in the company.
John: Correct. The revenue didn’t go up astronomically from the time we initially wanted to sell it. We just cut our spending, that’s all. Also at that time, CoCreate was quite attractive because it had a pretty large and loyal customer base. The revenue wasn’t growing much. But it wasn’t declining either. It was quite a stable revenue stream. So yes, we just made it profitable and sold it. But anyways, before the final deal, I had already left the company because I had got recruited to join another start up called Proficiency as CTO, which was a really interesting Israeli company. It had the idea of feature based interoperability. So what you did was you used the CAD system’s API to extract the features and their parameters and put into a kind of a standard intermediate format.
Deelip: So basically importing a feature based proprietary model from one MCAD system to another while maintaining the history tree.
John: Yes, it was kind of a cool idea and it worked pretty well. I think you could get about 90% of the features. But it turns out that its one of these things where 90% is not good enough, You need to have a 100% to make the system work. So the market for Proficiency software was very small. We tried to sell the company. But the problem was that none of the CAD vendors were interested because for our technology to work, you needed to be in the partner program of every CAD vendor. The moment one of them acquired us, we would be kicked out of the partner programs of everyone else. We tried repositioning it later as a complement to PLM systems and tried to get PLM people interested. But none of those strategies really worked. That company burned through a lot of money. They had a lot of venture capital, about 50 million I think and it never really produced anything close to that. Eventually, the US inventors got tired and backed out. It just didn’t have enough money to sustain itself and eventually they sold it to TrancenData.
Deelip: So where did you go after that?
John: Well, I did some consulting at various companies for about six months. Eventually I found a position at Microsoft to work on Citrix to work on terminal servers. I was at Citrix for only about a year before Keith Mountain contacted me saying he was looking for a VP or R&D for Spatial. Well, to me the IT business is pretty pedestrian as compared to CAD (laughs).
Deelip: Actually I find that amusing coming from a Cultural Anthropoligist.
John: (Laughs) Yeah. Anyways, I have been at Spatial for four years now.
Deelip: So what was one of the first things you did at Spatial?
John: Well, we started studying our customers and what they were doing. We wanted to know how long it took a customer to develop and actually ship a product based on ACIS. We found that 50% of time, they never made it because the project got canceled or something like that. As far as the rest we found that the average time to ship a product was three years. So this was like the orchard business where you don’t see the returns for a number of years (laughs). So we started to look at ways of changing things. Part of the problem was the sheer difficulty in learning how to use ACIS. I mean, no matter how much effort we put into our documentation, it still would take a lot of time for someone to learn how to build an application based on our components. So that’s where got the idea of building a platform for our technologies.
Deelip: You mean Rapid Application Development Framework (RADF).
John: Yes. We also figured that our customers would be interested in looking at other ways of delivering their software, maybe on the Cloud. So we decided that it would probably be a good idea to offer them RADF as a dual platform – for the desktop as well as the Cloud.
Deelip: John, it was really nice talking to you.
John: Same here.