A Conversation With Carl Bass – Part 1
Today at AU 2010 I sat down with Autodesk CEO Carl Bass and got his views on a few topics. We spoke about PLM, Dassault Systemes’ entry into the AEC, SolidWorks developing a SolidWorks killer, Infinite Computing and Autodesk’s entry into the Do It Yourself market.
Deelip: What exactly is your problem with PLM?
Carl: (Laughs) I don’t have a problem with PLM. I still stand by the comments I made around five years ago wherein I said that PLM was a technology in search of a problem. I think that’s changed. I think people have had a more pragmatic approach to PLM. You can now find working implementations of PLM systems. I think one of the problems with PLM when used in the press is that it is not a very clear term. People mean different things.
Deelip: I guess that’s because companies themselves are different and they have different needs that must be fulfilled by PLM.
Carl: True. But that’s not a reason for a term to mean different things. When you look at it, we actually have many things that people would call PLM offerings. Our Vault line of products do data and document management, check in, check out, revision, release to manufacturing, etc. I would suspect that we have way more implementations than anyone else out there in the market. A lot of my criticism was aimed at was the part where people were involved in these PLM implementations that cost them tens of millions of dollars. Customizations that never resulted in any value to the customer. Many of them failed. I mean, they got ECO’s to work. Is that worth millions of dollars? I think it was worthy being cynical and skeptical of the actual value to the customers. I think the market has changed somewhat and I do think customers are getting value from their PLM systems. Not always. Not from every vendor. But it is increasingly becoming a mainstream part of larger manufacturers.
Deelip: So from what I understand you don’t see much of a difference between PDM and PLM. Its just the management of product data. Do you consider the lifecycle part of it?
Carl: I don’t think I said that PLM is the wrong term. For example, you brought up lifecycle. If you look at document and data management it manages the lifecycle of the product. But if you want to get into things like say requirements, that’s a different part of the lifecycle in a different sense of the term. People need a much better definition of what PLM is before you can have an intelligent conversation about it. The problem with expanding the definition of PLM is that you can have someone say, “I’m building a system that helps marketing produce come material for sales in a manufacturing company“. That could be Creative Suite. So is Creative Suite now a PLM tool? I don’t think we can accept that any vendor can define PLM to be whatever they want it to be.
Deelip: I am assuming that you see Inventor being used in the Automotive and Aerospace industries.
Deelip: So don’t you think you need to change your messaging a little bit. I say that because those two industries have got around to accepting the term PLM and most of them already have PLM implementations in place. I mean, whether you like it or not, PLM is the term that is used by people in those industries. Since you have admitted that things have now changed, do you think its time for Autodesk to adjust its official stand on PLM?
Carl: We have made comments on PLM in the broad sense of the term. But like I said, I believe there are more people doing PLM with our software than any other vendor out there.
Deelip: But you are not marketing it as PLM.
Carl: True. And that’s because PLM is a confusing term. I think you are absolutely right when you say that in the Automotive and Aerospace industries PLM is an accepted term, even though it is vague, ambiguous and confusing. And they all have a notion of what it is for them. If you wrote an article on PLM and included Autodesk in it you certainly will not get a note from me saying, “Why did you include us?“. I’m not against it. I just think its not the best way to explain what it is that we do. Let me ask you something. Have you ever seen the word tablet used anywhere in Apple marketing? There is a category of hardware called tablets from different vendors.
Deelip: They don’t? I never noticed. Who don’t they?
Carl: Because they are selling something called the iPad. They are not interested in which category it fits into. The term tablet computing is limiting. It bunches stuff together that they don’t want to be a part of. But coming back to PLM, I am more than willing to admit that offerings in PLM have become better. They have become more common place. There are more successes now than there were before. And for many manufacturers its a critical part of their IT infrastructure.
Deelip: Switching gears a bit, I am sure you have an opinion on Dassault Systemes getting into AEC. My question is whether you would like to share it with me.
Carl: (Laughs) It’s interesting. Many of the manufacturing vendors have always talked about buildings. I’m sure their software is being used in construction somewhere. Regarding Dassault, for a number of years they had some interest in indirectly selling CATIA through Gehry Technolgies. They seemed to have abandoned that.
Deelip: But that as more on the specialist AEC side of things. But now they are getting into the core of where AutoCAD is being used. I am talking about DraftSight and Live Buildings. They are getting ready to step into your territory.
Carl: Really? I welcome them to compete. Like I said, they have made attempts to come into the AEC market. I don’t think they have the tools that are well suited. If you look back at Revit which is basically a parametric solid modeler. Even though many of the people who originally worked on it came from a manufacturing background, they decided to do it completely differently because they thought that manufacturing software was inappropriate for AEC. Many people think that you can take software made for manufacturing and turn it into something fit for AEC. And that’s not true. The other thing is, given the current problems that Dassault is having in manufacturing, I am actually curious why they are not focused on serving their customers in manufacturing. I don’t know what their fascination in AEC is.
Deelip: I guess its about expanding their markets because the manufacturing market isn’t growing all that much.
Carl: If you club software and services, the manufacturing market is probably between 10 and 15 billion dollars. The last time I checked, they couldn’t have been more than 15% of that market. So if you are one of the leaders in an industry and have only 15% market share, you should be thinking what you need to do to get to 18% or 20%. I’m not sure this is about expanding their markets. This is more like a distraction, given that they already have sizable issues with their ENOVIA products, V6 and its lack of adoption and this thing of combining ENOVIA with CATIA is not being well received by their customers. What they have told their SolidWorks customers about “end of lifing” their products is well… I have no idea why you would tell your customers that you are killing your product that has historically been successful in the market. If you look a the history for the first ten years they were a very successful, growing and profitable company. I think over the last handful of years, its fallen behind in terms of capability, it has lots of quality issues, a lack of resources have been directed towards SolidWorks…
Deelip: SolidWorks Labs is virtually dead.
Carl: Exactly. They have taken the resources and are directing it elsewhere. So now its got to this period of neglect. I would bring it back to success as opposed to going from success to neglect to death.
Deelip: Well, they are not going to just kill it. They are looking to replace it with a new product.
Carl: You can use whatever words you want. But the point is they are killing their old product and telling their customers that there is this new product that they would like them to switch to. And most of the new product seems to be not necessarily conceived with the best interests of the users. It seems to be directed by internal concerns. Things like, “How do we make our tools compatible with each other?“, “How do we save money by combining the engineering function so that we can use the same modeling engine?” I’m not sure many SolidWorks customers have raised their hand and said “I wish my data was interoperable with CATIA more” or “I wish you used the same modeling kernel as CATIA” or “I wish the UI on SolidWorks looked more like CATIA“. I hear them say, “Could you fix the bugs?“, “Could you make it more capable in this area or that?” I think those are more reasonable things for them to be focused on.
Deelip: What exactly is infinite computing and how it is different from cloud computing? What was the need to come up with this new term?
Carl: I think what we are doing is quite different than what most people have done on the cloud. Take SalesForce.com, one of the most commercially successful companies in cloud computing. The things that they do via the cloud is an appropriate application for a distributed sales force. Its just a a lightweight business application. We are thinking about a completely different application. Users will have options to use one CPU or ten thousand CPU’s. User’s of SalesForce.com have no need for that kind of option. We are not trying to launch a new term. What we are basically saying is that what we are interested in the cloud is quite different from what others are interested in. Its not like putting the software on more machines and solve the problem faster. This involves fundamentally changing the way we get solve problems.
Deelip: You mean something like Inventor Optimization Technology Preview.
Carl: Yes. Instead of a sequential iterative process of designing and testing over and over again, until I reach the thresh hold of acceptable or I run out of time, money or patience, let me talk to the computer at a higher level of abstraction and ask it to optimize along certain dimensions and give me the best answer. That’s a different way of solving a problem. So if people want to say SalesForce.com using the cloud to do business application is the same as what we are doing and call both cloud computing, I’m fine with that. I’m not going to complain. I just don’t think it is a very effective way of informing our customers what it is that we are doing. Having said that, we do have some pretty normal cloud applications such as Inventor Mobile and stuff like that.
Deelip: Buzz Kross made an interesting comment of taking the standalone Inventor Fusion application to the DIY market.
Carl: Yes, our job is to build tools. If the tools we build are appropriate in terms of their functionality and their price, they will find their way into the these different market segments. I have always been interested in that part of the market. In fact, I am part of that market myself. I am interested in people like students, all the way down to middle school. I’m interested in DIY. I think there are consumers who would like to do some sort of design and I am interested in giving them the tools.
>> Part 2