A Long Conversation With Mike Payne
I have been doing these interviews with captains of the CAD industry for a while now. I must admit, one of the people I wanted to interview the most was Mike Payne. For those who don’t know who Mike Payne is be sure to read the following conversation in its entirety and you will never forget who the man is. I guarantee it.
I have been bumping into Mike at the last three COFES events I attended. If you have ever spoken to him you will know that he minces no words when telling you exactly what he thinks about someone or something, as you will see in the following conversation. We spoke for more than an hour. So this is going to be a long one. I suggest that before you proceed you grab a coffee or whatever it is you drink when you take a long break.
I recorded the conversation on audio and then transcribed it. So rest assured you will get to hear Mike Payne in his purest form. I usually don’t do this. But given the content of the conversation I sent the transcribed conversation by email to Mike a while ago and asked him if he wanted to edit or remove something. He simply added the last names of a couple of people he mentioned in the conversation and sent it back to me.
So sit back and enjoy. You are going to remember this one.
=======
Deelip: I have heard a bunch of interesting Mike Payne stories from a lot of people. I’d like to know the Mike Payne story from the man himself.
Mike: (Laughs) Well, I went to Southampton University and did Electrical Engineering and then did a Masters in Physics with solid state vacuum devices. Later I got an MBA from Pace University in New York. Thereafter I was an adjunct professor in a couple of places.
Deelip: What did you teach?
Mike: At one of the New Jersey Institute of Technology I taught transmission lines and all that kind of stuff. I also did some Computer Science and Logic courses in the graduate college at Northeastern. That was a long time ago, around the late seventies or early eighties.
Deelip: These names are American. You are British, right? When did you move to the US?
Mike: Oh! In 1969 the company that I worked for sent me to Philadelphia for a year. I didn’t like it. So I left.
Deelip: You didn’t like the company or the work?
Mike: I didn’t like the company. They were rogues. Anyways, I then got a job at RCA where I designed chips and tested them. I started out designing chips for the RCA computer division. Then I worked on television stuff. Then they got this contract to build the logic circuits for the Trident missile program. And so somehow I got the run that team. This was from 1971 to about 1975. Then I went to a company in Massachusetts which probably was a mistake and then I went to Prime Computer. I tried to bring chip technology to them but never succeeded because they were so ingrained in the previous way of doing it with tubes. They could not imagine that everything could be reduced to a chip. That was a disaster and they got bought by Computervision.
Deelip: If I remember correctly you met Keith Mountain there the first time. And you and he were trying to sell Prime the idea of getting into CAD.
Mike: Absolutely. The thing with chips was that you need to have an environment to test them. You couldn’t really test them in the real world because they were on the end of a missile. So you had to do a lot of simulation. They had a program but it wasn’t capable of doing it. So I went to the CAD people in their nice little empire and asked them if they could modify their program to take into account certain things that I really can’t talk about here because it’s classified information. They said that it was not on their schedule. So I told them that I would do it and they started laughing. But I got access to their software and then I did it. This was at the time when a computer was a room. An IBM 370 or something like that. So I went back to them and told them it’s done and they stopped laughing. Prime needed the same software to make chips and simulate things. Someone got this bright idea to get rights to CAD software, access to PDMS, which is a plant software from the CADCentre in Cambridge and partial rights to MEDUSA and put it along with a home grown electronic product for doing these logic boards and create a CAD division, if you will.
Deelip: So were you involved in starting that?
Mike: No, let’s just say it was the CFO. So the idea was that if you had the application then you might sell more computers. So I ran the development of that and I had enough of it. So I left. I had previously hired somebody from Computervision and I got a call from his brother Sam Geisberg saying, “Why don’t you come and have breakfast with me and I’ll show you something?“. So I went and met Sam and Danny Dean. He showed me a bracket with a hole that he could move and I said, “Yeah, I think we can do something with this, Sam.” And that’s how PTC started.
Deelip: So Sam invented this idea of history based parametric modeling.
Mike: Yeah, it was Sam’s idea and Danny was doing the coding and I joined him. Later we got some money, which is another long story. And then we got this guy Leonid, another Russian, who did the geometry.
Deelip: So were you the CTO at PTC?
Mike: Well, I was one of the development managers. I was there for about eight years and I left. Then I got asked to do some due diligence by a venture guy on something Jon Hirschtick was doing. After I looked at it and gave my opinion the venture guy said that he would give the money if I joined in. I got released from a non-compete from PTC and that’s how SolidWorks started.
Deelip: So would it be fair to say that SolidWorks got the money to start off because of your links to PTC?
Mike: Well, not so much the link to PTC. But I would say that the venture guy decided that the track record was good enough for him. At that time I promised the VC’s that we would have a product out by AutoFact ’95, a trade show that existed back then. And they didn’t believe me, but we did it. I was the VP of R&D at SolidWorks.
Deelip: You mentioned in your keynote today that you chose Parasolid over ACIS for SolidWorks. Would like you elaborate a little on that?
Mike: (Laughs) You really want the story. OK, here’s how it went. The original prototype was built using ACIS. Jon Hirschtick had talked a number of component vendors like D-Cubed to give him their stuff and he would pay for them later. One of them was ACIS and so they built the prototype with ACIS. So I now get involved and have a look at it and say “but we can’t make this and we can’t make that“. Scott Harris had a set of rounds that he wanted to work to be sure that he had a robust product. They had worked fine in Pro/ENGINEER but most of them failed in ACIS. So we met with Spatial and told them that their product had some problems. And I didn’t like the answers that I got. So then we decided to get a hold of Parasolid and put that in. So I got someone to put Parasolid in and all of a sudden most of these damn rounds started working. At the same time, I called Dick Harrison, the then VP of Sales at PTC and asked him to meet with us. So I took him through the business plan and we gave him a demo. He asked me, “So what do you want from me?“. I said, “Dick, you got the best geometry and I want to license it.” This was while we were trying to put in Parasolid. He said, “But we could end up as competitors“. I said, “Look, I have laid out our business plan. Our goal is to go after the 2D guys and give them 3D.” But he didn’t like the idea. He also told me that if we ended up as competitors he would demand a benchmark. As it turns out the first time we did run into Pro/E later on, he demanded a benchmark and he lost. The same thing happened again and they never asked for another benchmark.
Deelip: But you said that your business plan was to take on the AutoCAD users, not Pro/E.
Mike: I know. But then we did end up fighting each other. The thing is the VAR’s. We didn’t want them to go after Pro/E users. We wanted them to go after the 2D guys. But you know VAR’s. They go after the low hanging fruit. They couldn’t sell something to Pro/E users before. So they went ahead and started selling them SolidWorks. It then became obvious to me that we needed to have good modeling and add more features because I knew that PTC would be gunning for us. So I hired a guy who could do surfacing. The enemy was AutoCAD but we finally ended up fighting Pro/ENGINEER. This surely was a kick in the posterior for SolidWorks because people started asking for all kinds of things and expected us to deliver. It sure couldn’t do everything that Pro/E could, but it could do a lot.
Deelip: So at that time was Solid Edge around?
Mike: Well, that’s another problem. UG started developing Solid Edge in Huntsville six months earlier to us developing SolidWorks and we didn’t know about it.
Deelip: Hold on a second. Are you saying that UG agreed to license Parasolid to you for SolidWorks and you didn’t know that they were developing a rival product using the same modeling kernel?
Mike: Correct. We had no idea about SolidEdge at that time. So they came up with this Jupiter Technology and they took ages doing it. They eventually came out with a product with less capability than the initial release of SolidWorks but about six months after. So they took about twice the time to develop it as us.
Deelip: And where was Inventor at that time?
Mike: Inventor was nowhere in sight. Prior to coming out at AutoFact Autodesk had been trying to come out with a 3D product. And somehow they came and talked to us. Eventually they actually tried to buy the company.
Deelip: You mean buy SolidWorks before you had shipped the product?
Mike: Correct. They offered a certain amount of money which was about 50 cents a share less than what the VC’s would have agreed to, a nice return actually. But they made it too rigid. They said that we would have to use the AutoCAD database. And so we said no.
Deelip: That doesn’t make sense.
Mike: I know that it doesn’t make sense. But that’s what they demanded.
Deelip: Who was heading Autodesk at that time?
Mike: Dominic Gallelo. He heads MSC today. He was in charge of the MCAD side of Autodesk’s business, like Buzz Kross is today. So he decided that if he was going to spend X million dollars buying a company, they could rather invest that money. So they produced Inventor which was not as good as SolidWorks and it still isn’t.
Deelip: Why? Because it was running on ACIS at that time?
Mike: No, not because of that. It’s just not good enough. Pro/E is probably still the best feature based modeler, certainly in terms of geometry. It’s got an awful UI. The company has upset customers. But it still is a decent product. Solid Edge is a better product than Inventor. But it had sold, if you are lucky, a third of the seats of SolidWorks. It just hadn’t been sold very well. Basically, they couldn’t get out of their own way. Now Inventor has the home team advantage. They got a huge customer base, but that doesn’t make it a better product. It comes back to, you know “It’s the economy stupid“. Well in this game “It’s the geometry stupid“. And they are not very good at making the geometry. If you look at the mechanical things around here, not all of them are like airplane wings. Most of the objects are like that lamp over there (points to a simple looking lamp). Am I trying to suggest that you can’t make that in Inventor? I wouldn’t be so silly. But take this lamp for example (points to another curvy lamp). Try and make that in Inventor. In Mechanical there is this range of complexity of shapes and an OK product could probably do half of them.
Deelip: I am trying to get a sense of how you compare Inventor with SolidWorks at that time. Was it something like how Alibre Design is being compared with SolidWorks?
Deelip: Alibre? Well, it’s a not very capable product for a not very capable price. And I don’t get this underserved market nonsense. If you are making something, you got to be able to make it, OK? No, there is no comparison between Alibre and Inventor. I may have criticized Inventor a bit. But I mean, this is like night and day.
Deelip: So for how long did you run the development of SolidWorks?
Mike: For about five years. In the middle it got sold to Dassault Systemes.
Deelip: What exactly went on during that time of selling it?
Mike: Well, actually we were interested in going public at that time. At one of the AutoFact’s I met this guy called Phillipe Herbert. He was doing the business development for Dassault Systemes. So I got an invitation with Scott Harris to go to Paris under the banner of cooperation. They didn’t have a product on Windows. They were developing CATIA V5 at that time and IBM was complaining that they couldn’t sell to the Windows people. So we had various meetings with people to discuss this and that and there was a certain amount of transparency. On our side, I must add. Then I get invited to this lunch at the executive dining room of Dassault Aviation where I have lunch with Bernard Charles and Thibault de Tersant, who I had never met before. There we talk about how we could cooperate, exchange of data and all that stuff. They couldn’t stand PTC. I remember at one time Bernard started complaining to me about PTC. He said, “You know, PTC bought a CATIA license and they started filing bug reports“. And I said to Bernard, “That was me, Bernard.” (Laughs) They changed the subject. Anyways, at one point the conversation Thibault said, “But we could always buy you“. I looked at him as said, “We are not planning on that. We are thinking of going public. In any case, we are nice girls. But we are not a cheap date.” So we go back to Boston and nothing happens for a while. Then we get a call, then we get an offer and we say no. This goes on for a while and finally we get an offer that the VC’s liked and that’s how Dassualt Systemes bought us. This was in 1997.
Deelip: And you stayed on.
Mike: Well that was the deal. We basically shook hands for staying for three years and we did. I was doing the same job as I was doing at PTC and I knew the next chapter in the book. I mean I’ve been there, done that and now it’s no fun. So that’s how I ended up going to Paris as CTO of Dassault Systemes. I did a few things there and all of a sudden Bernard tells me that they had decided to buy Spatial. And I said, “That’s a bit of a problem, Because I’ve have some history with them.” I knew things weren’t good out there. He continued, “And I’d like you to run it“. I told him that I would think about it. I couldn’t think of anyone that could fix it. So I agreed to do it myself.
Deelip: So how was it at Spatial?
Mike: Well, everything had to be fixed. It was worse than I thought. The product was unreliable. It had half-finished capabilities. Customers were mad. Customer service was non existent. The pricing was all over the place depending on what somebody would be able to pay at the end of a quarter. And there was zero quality assurance.
Deelip: I guess that was about the time Autodesk was making a noise of how the bugs in ACIS were not being fixed.
Mike: Oh yes. It was every bit as bad as they said. They and a few other customers. The bugs didn’t get fixed. Memory leaked like nobody’s business. Then Buzz Kross comes out to see me and here is where he made a big mistake. I told him, “I’ve looked into this. You are completely right. I’m gonna fix it. I can’t fix it all at once. But that’s why I’m here.” And he didn’t believe me.
Deelip: I believe this was ACIS V7. That’s when Autodesk exercised their option to buy the ACIS source code, right? That’s why I guess all Autodesk software do not output ACIS SAT files later then V7.
Mike: Yeah, right. I remember. So we made some major strides in V8. Didn’t fix everything, obviously. We fixed the memory leaks. I remember, the tolerant modeling was only on certain operators. It was for demonstration and then they put it in the product. I mean, of whats use of that? You might as well not have it. I instituted quality control. That pissed some people at Spatial off. I mean, I bought computers and ran tests (shrugs shoulders). I fixed the contracts and the pricing. They were all different. That’s crazy.
Deelip: So each company had different contracts?
Mike: Yeah, it was like who could you rip off for the most. It does not work that way in the components business. All customers need to be treated the same. I killed the notion of different pricing for different customers. Slowly but surely we got everything on track and we got it to be profitable, after which I left. That was in 2005.
Deelip: And then you didn’t stay idle for long, now did you?
Mike: Well, I found something else to do. I joined forces with David Taylor and Blake Courter and we did SpaceClaim. Around 2004 Blake had been bugging me when I was at Spatial. I orchestrated a meeting with them and Dassault Systemes in Paris. They showed the prototype of SpaceClaim that they had put together. Bernard liked it but nothing happened. Later in the year I started giving Blake some advice on how to deal with the venture guys and raise money. But to cut a long story short, they ran into a brick wall. So eventually they asked me to join. I got Danny Dean to join in as well and I raised money pretty quickly. Dassault Systemes still had some interest. They wanted to invest in it. But they wanted too big a piece of it.
Deelip: So was the idea of SpaceClaim to get it up to a point where it could be sold or was it created to be run?
Mike: No, you worry about that later. The first thing in a start-up company is you want to get to cash flow positive. Then you start worrying about what happens later. It’s like SolidWorks. We had never planned that we were going to get bought by Dassault Systemes. In the case of PTC, we thought we would get bought by somebody. But we didn’t. So we went public. We had conversations with Computervision at one point. It turns out it ended up being the other way around. You certainly don’t plan these things. Once you are cash flow positive, you got some options.
Deelip: So what’s so special about SpaceClaim, the software?
Mike: You remember what I put on those slides today about computers getting faster? This why I believe CoCreate really isn’t… And I feel the same way about KeyCreator. SpaceClaim was designed for these computers. It’s like how Microsoft consumes CPU cycles for their fancy UI. SpaceClaim consumes them to do things that the user doesn’t know how to do, without getting into details here.
Deelip: So you mean, instead of making the user smart, you are making the software smart.
Mike: Yes, and by the way, you see that with a lot of computer software. The user is required to be more average.
Deelip: When I recently spoke to Blake, he told me that they had stopped acting like a start-up company.
Mike: Yeah, there’s an element of truth in that. In SpaceClaim you can now build in a conceptual way many types of shapes. Can you build and manipulate every object? I wouldn’t make that claim. And by the way, I can’t make every object in CATIA either. That’s why they have development teams. So as each rev goes by, you add to that capability. There is tremendous sheet metal capability in SpaceClaim. But there’s more work to be done.
Deelip: No, I think meant that from a business standpoint.
Mike: Oh, I am discussing the product and the rest flows. In a start-up company you are struggling to add features. Now it’s more of broadening the product with things like sheet metal. You see, the paradigm of Direct Modeling lends itself to the concept of Sheet Metal much more than the feature based stuff. So when you get to that point you are no longer a start-up company from the development point of view. And the same thing happens in the sales and marketing. Its more planned than opportunistic. I mean the Tyco deal took a long time to put together. But then there was a plan that needed to be laid out for that to happen.
Deelip: So what is your role now in the company?
Mike: I’m just the Chairman of the Board. If they ask me to do something I do it. I don’t run the company.
Deelip: So what’s next?
Mike: Well, I’ve been advising an Israeli VC on something among other things. I’m looking for my next gig. But I don’t know what it is yet.
Deelip: How old are you?
Mike: 66.
Deelip: You are 66 and are looking for your next gig. So I take it you don’t plan on retiring anytime soon.
Mike: Retiring? Do you know what a rocking chair looks like? Some people think that’s a great idea. I mean, they say you can now do this and that. They never do this and that. You need to keep you mind active. Have challenges.
Deelip: I have heard a bunch of interesting Mike Payne stories from a lot of people. I’d like to know the Mike Payne story from the man himself.
Mike: (Laughs) Well, I went to Southampton University and did Electrical Engineering and then did a Masters in Physics with solid state vacuum devices. Later I got an MBA from Pace University in New York.
Thereafter I was an adjunct professor in a couple of places.
Deelip: What did you teach?
Mike: At one of the New Jersey Institute of Technology I taught transmission lines and all that kind of stuff. I also did some Computer Science and Logic courses in the graduate college at Northeastern. That was a long time ago, around the late seventies or early eighties.
Deelip: These names are American. You are British, right? When did you move to the US?
Mike: Oh! In 1969 the company that I worked for sent me to Philadelphia for a year. I didn’t like it. So I left.
Deelip: You didn’t like the company or the work?
Mike: I didn’t like the company. They were rogues. Anyways, I then got a job at RCA where I designed chips and tested them. I started out designing chips for the RCA computer division. Then I worked on television stuff. Then they got this contract to build the logic circuits for the Trident missile program. And so somehow I got the run that team. This was from 1971 to about 1975. Then I went to a company in Massachusetts which probably was a mistake and then I went to Prime Computer. I tried to bring chip technology to them but never succeeded because they were so ingrained in the previous way of doing it with tubes. They could not imagine that everything could be reduced to a chip. That was a disaster and they got bought by Computervision.
Deelip: If I remember correctly you met Keith Mountain there the first time.
And you and he were trying to sell Prime the idea of getting into CAD.
Mike: Absolutely. The thing with chips was that you need to have an environment to test them. You couldn’t really test them in the real world because they were on the end of a missile. So you had to do a lot of simulation. They had a program but it wasn’t capable of doing it. So I went to the CAD people in their nice little empire and asked them if they could modify their program to take into account certain things that I really can’t talk about here because it’s classified information. They said that it was not on their schedule. So I told them that I would do it and they started laughing. But I got access to their software and then I did it. This was at the time when a computer was a room. An IBM 370 or something like that. So I went back to them and told them it’s done and they stopped laughing. Prime needed the same software to make chips and simulate things. Someone got this bright idea to get rights to CAD software, access to PDMS, which is a plant software from the CADCentre in Cambridge and partial rights to MEDUSA and put it along with a home grown electronic product for doing these logic boards and create a CAD division, if you will.
Deelip: So were you involved in starting that?
Mike: No, let’s just say it was the CFO. So the idea was that if you had the application then you might sell more computers. So I ran the development of that and I had enough of it. So I left. I had previously hired somebody from Computervision and I got a call from his brother Sam Geisberg saying, “Why don’t you come and have breakfast with me and I’ll show you something?”. So I went and met Sam and Danny Dean. He showed me a bracket with a hole that he could move and I said, “Yeah, I think we can do something with this, Sam.” And that’s how PTC started.
Deelip: So Sam invented this idea of history based parametric modeling.
Mike: Yeah, it was Sam’s idea and Danny was doing the coding and I joined him. Later we got some money, which is another long story. And then we got this guy Leonid, another Russian, who did the geometry.
Deelip: So were you the CTO at PTC?
Mike: Well, I was one of the development managers. I was there for about eight years and I left. Then I got asked to do some due diligence by a venture guy on something Jon Hirschtick was doing. After I looked at it and gave my opinion the venture guy said that he would give the money if I joined in. I got released from a non-compete from PTC and that’s how SolidWorks started.
Deelip: So would it be fair to say that SolidWorks got the money to start off because of your links to PTC?
Mike: Well, not so much the link to PTC. But I would say that the venture guy decided that the track record was good enough for him. At that time I promised the VC’s that we would have a product out by AutoFact ’95, a trade show that existed back then. And they didn’t believe me, but we did it. I was the VP of R&D at SolidWorks.
Deelip: You mentioned in your keynote today that you chose Parasolid over ACIS for SolidWorks. Would like you elaborate a little on that?
Mike: (Laughs) You really want the story. OK, here’s how it went. The original prototype was built using ACIS. Jon Hirschtick had talked a number of component vendors like D-Cubed to give him their stuff and he would pay for them later. One of them was ACIS and so they built the prototype with ACIS. So I now get involved and have a look at it and say “but we can’t make this and we can’t make that”. Scott Harris had a set of rounds that he wanted to work to be sure that he had a robust product. They had worked fine in Pro/ENGINEER but most of them failed in ACIS. So we met with Spatial and told them that their product had some problems. And I didn’t like the answers that I got. So then we decides to get a hold of Parasolid and put that in. I got someone to put Parasolid in and all of a sudden most of these damn rounds started working. At the same time, I called Dick Harrison, the then VP of Sales at PTC and asked him to meet with us. So I took him through the business plan and we gave him a demo. So he asked me, “So what do you want from me?”. I said, “Dick, you got the best geometry and I want to license it.” This was while we were trying to put in Parasolid. He said, “But we could end up as competitors”. I said, “Look, I have laid out our business plan. Our goal is to go after the 2D guys and give them 3D.” But he didn’t like the idea. He also told me that if we ended up competitors he would demand a benchmark. As it turns out the first time we did run into Pro/E later on, he demanded a benchmark and he lost. The same thing happened again and they never asked for another benchmark.
Deelip: But you said that your business plan was to take on the AutoCAD users, not Pro/E.
Mike: I know. But then we did end up fighting each other. The thing is the VAR’s. We didn’t want them to go after Pro/E users. We wanted them to go after the 2D guys. But you know VAR’s. They go after the low hanging fruit.
They couldn’t sell something to Pro/E users before. So they went ahead and started selling them SolidWorks. It then became obvious to me that we need to have good modeling and add more features because I know that PTC would be gunning for us. So I hired a guy who could do surfacing. The enemy was AutoCAD but we finally ended up fighting Pro/ENGINEER. This surely was a kick in the posterior for SolidWorks because people started asking for all kinds of things and expected us to deliver. It sure couldn’t do everything that Pro/E could, but it could do a lot.
Deelip: So at that time was Solid Edge around?
Mike: Well, that’s another problem. They started developing Solid Edge in Huntsville six months earlier to us developing SolidWorks and we didn’t know about it.
Deelip: Hold on a second. Are you saying that UG agreed to license Parasolid to you for SolidWorks and you didn’t know that they were developing a rival product using the same modeling kernel?
Mike: Correct. We had no idea about SolidEdge at that time. So they came up with this Jupiter Technology and they took ages doing it. They eventually came out with a product with less capability than the initial release of SolidWorks but about six months after. So they took about twice the time to develop it as us.
Deelip: And where was Inventor at that time?
Mike: Inventor was nowhere in sight. Prior to coming out at AutoFact Autodesk had been trying to come out with a 3D product. And somehow they came and talked to us. Eventually they actually tried to buy the company.
Deelip: You mean buy SolidWorks before you had shipped the product?
Mike: Correct. They offered a certain amount of money which was about 50 cents a share less than what the VC’s would have agreed to, a nice return actually. But they made it too rigid. They said that we would have to use the AutoCAD database. And so we said no.
Deelip: That doesn’t make sense.
Mike: I know that it doesn’t make sense. But that’s what they demanded.
Deelip: Who was heading Autodesk at that time.
Mike: Dominic Gallelo. He heads MSC today. He was in charge of the MCAD side of Autodesk’s business, like Buzz Kross is today. So he decided that if he was going to spend X million dollars buying a company, they could rather invest that money. So they produced Inventor which was not as good as SolidWorks and it still isn’t.
Deelip: Why? Because it was running on ACIS at that time?
Mike: No, not because of that. It’s just not good enough. Pro/E is probably still the best feature based modeler, certainly in terms of geometry. It’s got an awful UI. The company has upset customers. But it still is a decent product. Solid Edge is a better product than Inventor. But it had sold, if you are lucky, a third of the seats of SolidWorks. It just hadn’t been sold very well. Basically, they couldn’t get out of their own way. Now Inventor has the home team advantage. They got a huge customer base, but that doesn’t make it a better product. It comes back to, you know “It’s the economy stupid?”. Well in this game “it’s the geometry stupid”, and they are not very good at making the geometry. If you look at the mechanical things around here, not all of them are like airplane wings. Most of the objects are like that lamp over there (points to a simple looking lamp). Am I trying to suggest that you can’t make that in Inventor? I wouldn’t be so silly. But take this lamp for example (points to another curvy lamp). Try and make that in Inventor. In Mechanical there is this range of complexity of shapes and an OK product could probably do half of them.
Deelip: I am trying to get a sense of how you compare Inventor with SolidWorks at that time. Was it something like how Alibre Design can now be compared with SolidWorks?
Deelip: Alibre? Well, it’s a not very capable product for a not very capable price. And I don’t get this underserved market nonsense. If you are making something, you got to be able to make it, OK? No, there is no comparison between Alibre and Inventor. I may have criticized Inventor a bit. But I mean, this is like night and day.
Deelip: So for how long did you run the development of SolidWorks?
Mike: For about five years. In the middle it got sold to Dassault Systemes.
Deelip: What exactly went on during that time of selling it?
Mike: Well, actually we were interested in going public at that time. At one of the AutoFact’s I met this guy called Phillipe Herbert. He was doing the business development for Dassault Systemes. So I got an invitation with Scott Harris to go to Paris under the banner of cooperation. They didn’t have a product on Windows. They were developing CATIA V5 at that time and IBM was complaining that they couldn’t sell to the Windows people. So we had various meetings with people to discuss this and that and there was a certain amount of transparency. On our side, I must add. Then I get invited to this lunch at the executive dining room of Dassault Aviation where I have lunch with Bernard Charles and Thibault de Tersant, who I had never met before. There we talk about how we could cooperate, exchange of data and all that stuff. They couldn’t stand PTC. I remember at one time Bernard started complaining to me about PTC. He said, “You know, PTC bought a CATIA license and they started filing bug reports”. And I said to Bernard, “That was me, Bernard.” (Laughs) They changed the subject. Anyways, at one point the conversation Thibault said, “But we could always buy you”. I looked at him as said, “We are not planning on that. We are thinking of going public. In any case, we are nice girls. But we are not a cheap date.” So we go back to Boston and nothing happens for a while. Then we get a call, then we get an offer and we say no. This goes on for a while and finally we get an offer that the VC’s liked and that’s how Dassualt Ssytemes bought us. This was in 1997.
Deelip: And you stayed on.
Mike: Well that was the deal. We basically shook hands for staying for three years and we did. I was doing the same job as I was doing at PTC and I knew the next chapter in the book. I mean I’ve been there, done that and now it’s no fun. So that’s how I ended up going to Paris as CTO of Dassault Systemes.
I did a few things there and all of a sudden Bernard tells me that they had decided to buy Spatial. And I said, “That’s a bit of a problem, Because I’ve have some history with them.” I knew things weren’t good out there. He continued, “And I’d like you to run it”. I told him that I would think about it. I couldn’t think of anyone that could fix it. So I agreed to do it myself.
Deelip: So how was it at Spatial?
Mike: Well, everything had to be fixed. It was worse than I thought. The product was unreliable. It had half-finished capabilities. Customers were mad. Customer service was not existent. The pricing was all over the place depending on what somebody would be able to pay at the end of a quarter. And there was zero quality assurance.
Deelip: I guess that was about the time Autodesk was making a noise of how the bugs in ACIS were not being fixed.
Mike: Oh yes. It was every bit as bad as they said. They and a few other customers. The bugs didn’t get fixed. Memory leaked like nobody’s business.
Then Buzz Kross comes out to see me and here is where he made a big mistake.
I told him, “I’ve looked into this. You are completely right. I’m gonna fix it. I can’t fix it all at once. But that’s why I’m here.” And he didn’t believe me.
Deelip: I believe this was ACIS V7. That’s when Autodesk exercised their option to buy the ACIS source code, right? That’s why until now all Autodesk software do not output ACIS SAT files later then V7.
Mike: Yeah, right. I remember. So we made some major strides in V8. Didn’t fix everything, obviously. We fixed the memory leaks. I remember, the tolerant modeling was only on certain operators. It was for demonstration and then they put it in the product. I mean, of what use of that? You might as well not have it. I instituted quality control. That pissed some people at Spatial off. I mean, I bought computers and ran tests (shrugs shoulders).
I fixed the contracts and the pricing. They were all different. That’s crazy.
Deelip: So each company had different contracts?
Mike: Yeah, it was like who could you rip off for the most. It does not work that way in the components business. All customers need to be treated the same. I killed the notion of different pricing for different customers.
Slowly but surely we got everything on track and we got it to be profitable, after which I left. That was in 2005.
Deelip: And then you didn’t stay idle for long, now did you?
Mike: Well, I found something else to do. I joined forces with David Taylor and Blake Courter and we did SpaceClaim. Around 2004 Blake had been bugging me when I was at Spatial. I orchestrated a meeting with them and Dassault Systemes in Paris. They showed the prototype that they had put together.
Bernard liked it but nothing happened. Later in the year I started giving Blake some advice on how to deal with the venture guys and raise money. But to cut a long story short, they ran into a brick wall. So eventually they asked me to join. I got Danny Dean to join in as well and I raised money pretty quickly. Dassault Systemes still had some interest. They wanted to invest in it. But they wanted too big a piece of it.
Deelip: So was the idea of SpaceClaim to get it up to a point where it could be sold or was it created to be run?
Mike: No, you worry about that later. The first thing in a start-up company is you want to get to cash flow positive. Then you start worrying about what happens later. It’s like SolidWorks. We had never planned that we were going to get bought by Dassault Systemes. In the case of PTC, we thought we would get bought by somebody. But we didn’t. So we went public. We had conversations with Computervision at one point. It turns out it ended up being the other way around. You certainly don’t plan these things. Once you are cash flow positive, you got some options.
Deelip: So what’s so special about SpaceClaim, the software?
Mike: You remember what I put on those slides today about computers getting faster? This why I believe CoCreate really isn’t… And I feel the same way about KeyCreator. SpaceClaim was designed for these computers. It’s like how Microsoft consumes CPU cycles for their fancy UI. SpaceClaim consumes them to do things that the user doesn’t know how to do, without getting into details.
Deelip: So you mean, instead of making the user smart, you are making the software smart.
Mike: Yes, and by the way, you see that with a lot of computer software. The user is required to be more average.
Deelip: When I recently spoke to Blake, he told me that they had stopped acting like a start-up company.
Mike: Yeah, there’s an element of truth in that. In SpaceClaim you can now build in a conceptual way many types of shapes. Can you build and manipulate every object? I wouldn’t make that claim. And by the way, I can’t make every object in CATIA either. That’s why they have development teams. So as each rev goes by, you add to that capability. There is tremendous sheet metal capability in SpaceClaim. But there’s more work to be done.
Deelip: No, I think meant that from a business standpoint.
Mike: Oh, I am discussing the product and the rest flows. In a start-up company you are struggling to add features. Now it’s more of broadening the product with things like sheet metal. You see, the paradigm of Direct Modeling lends itself more to the concept of Sheet Metal much more than the feature based stuff. So when you get to that point you are no longer a start-up company from the development point of view. And the same thing happens in the sales and marketing. Its more planned than opportunistic. I mean the Tyco deal took a long time to put together. But then there was a plan that needed to be laid out for that to happen.
Deelip: So what is your role now in the company?
Mike: I’m just the Chairman of the Board. If they ask me to do something I do it. I don’t run the company.
Deelip: So what’s next?
Mike: Well, I’ve been advising an Israeli VC on something among other things. I’m looking for my next gig. But I don’t know what it is yet.
Deelip: How old are you?
Mike: 66
Deelip: You are 66 and are looking for your next gig. So I take it you don’t plan on retiring anything soon.
Mike: Retiring? Do you know what a rocking chair looks like? Some people think that’s a great idea. I mean, they say you can now do this and that.
They never do this and that. You need to keep you mind active. Have challenges.