Autodesk product development
23 September 2009
Martyn Day, talked with Autodesk CEO, Carl Bass about the company’s products - past, present and future.
Once upon a time, there was a group which called itself ‘Marin Software Partners’, set up by a handful of programmers, which developed some products for the emerging IBM compatible market. There was a CAD program called MicroCAD, a spreadsheet called OptiCalc and for filing and storing files was one called Autodesk. Eventually the company came to call itself Autodesk and its innovative 2D MicroCAD offering became AutoCAD, which funded the rapid growth of the largest design software company on the planet.
In the last 27 years, Autodesk has been through many changes and challenges, the most recent of which has been breaking the perceptions that the company only has one product. For many ‘Autodesk’ watchers, it’s clear the company has gone through several key stages of development, some being more innovative than others. Autodesk’s formidable dealer channel and marketing prowess have dominated the 2D design market from the early days of 8-bit processors, 640k RAM, DOS and the IBM PC - all the way through to Windows Vista, 64-bit processors and Gigabytes with everything. Looking at the history of computing, it truly has been a long and successful journey for Autodesk, as there aren’t that many software or hardware companies that have managed to stay around to tell the tale.
Looking at the recent innovative and prolific output from the company and remembering what seemed like decades of small enhancements to dumb 2D tools, I wondered what happened to Autodesk’s old view on product development? The company has a relatively new Chief Executive Officer (CEO), Carl Bass, and in our recent communications, it seemed worthwhile to ask questions charting how Autodesk got to where it is, what’s happening now and where is it all going?
Carl Bass originally co-founded a company called Ithaca Software, a developer of 3D tools, which was acquired by Autodesk in 1993. Bass formerly held a number of posts within Autodesk, including Chief Operating Officer (COO), Chief Technical Officer (CTO) and Vice President of the AEC division. Prior to that, Bass had previously left the company to head-up an Autodesk-funded ‘dotcom’ spin-off and even managed to drop out of the company once after being ‘fired’ for a couple of months by the previous CEO, Carol Bartz, for being a ‘pain in the ass’. Bartz and Bass eventually formed a good working relationship and for the five years before becoming CEO in 2006, took a leading role in planning the strategy of the company, eventually leading to a complete hand over of power when Bartz left. Under Bass, Autodesk has certainly undergone a product development renaissance, even though the route to market and business methodology hasn’t changed too drastically. As a side note, Carol Bartz has subsequently got the job of attempting to turn Yahoo! around.
If I were to appear on Mastermind, I guess one of my specialist topics could be the history of AutoCAD. I’ve been an AutoCAD drafter and a CAD journalist for err, ummm, a couple of decades, and I’ve used every release since version 9 of AutoCAD. It seemed to me that for the longest time, Autodesk gave the appearance of being a one product company, as AutoCAD was used as a ‘single remedy’ solution. Any new market would be addressed with a new flavour or add-on to AutoCAD and there were precious few new code streams coming from the company. You could have the answer to any problem, so long as it was AutoCAD.
Bass explained, “It’s really not quite that simple. Nineteen years ago we had a substantially different group of products. We always had many different platforms, some have made it, some have not. Back in those days, we had other code streams such as Hyperchem, Animator Studio, Generic CADD, AutoSketch, 3D Studio and Xanadu, a number of which had long lives. The broad application of AutoCAD came about as we noticed that AutoCAD was being used in multiple markets and from that we enhanced it to better address key vertical markets.”
With the growing penetration into Mechanical and AEC firms, Bartz created market divisions within Autodesk to develop applications that provided bespoke flavours of AutoCAD and enhanced productivity, resulting in releases such as AutoCAD AEC (later Architectural Desktop), AutoCAD Map and AutoCAD Mechanical. To speed this process up, Autodesk started to acquire its biggest third party developers such as Genius, a mechanical developer from Germany and Softdesk, an AEC developer from the United States.
At this time, increasing demands were placed on AutoCAD’s 2D feature set and existing software architecture. By all accounts the move from AutoCAD R11 to R12 was a massive task for the then CEO John Lynch. AutoCAD was showing its programmatic age and was one giant monolithic lump of code. The decision was made to re-architect around the modular, Object Oriented, C++ language and to add a 3D kernel and target the next release as being the first proper Windows version. Essentially the company tried to do too much in one release and not only did it re-architect the foundation of AutoCAD but the development team also built on top of those new features, adding great technology such as the ACIS 3D engine. R13 shipped too soon, lacked speed, ate hardware and lacked quality. Microsoft compounded matters by delaying the release of the target operating system, Windows NT. It was a multi-year set back for the company. Lessons were learnt and there were many positive impacts on the way software was developed and quality tested in all subsequent product releases. The move to Windows and Objects may have had a shaky start but the company survived the crossing.
I asked Bass what were the legacies that Bartz left the company. He replied, “Three things happened under Carol. The company made that technology transition from DOS to Windows and with probably the exception of Microsoft, very few companies that were industry leaders on DOS managed to remain industry leaders when on Windows. She was also instrumental in driving the company into market groups. And finally, I think people forget just how young and immature Autodesk was back then when she took over. So there was a task to make it a more mature and professional company. Autodesk was a company that had a small number of employees and had taken off like a rocket ship and needed some firm guidance.”
In terms of any difference between Autodesk under Bartz vs Bass, “It’s hard to demarcate because there was such a long period of handing over the reigns”, replied Bass. “Since Carol, there has been a lot of product development and broadening of our product’s capabilities and an increase in the number of platforms supported. Take products like AutoCAD and LT, even though they are many years ‘old’ and have the same brand names, they can do so much more than they could 15 years ago.”
Bass highlighted how the design technology has changed in the time Autodesk has existed, “Nowadays we really are creating digital models that are facsimiles of what we create in the real world. We started out with lines that represented something, dashes and dots and thicknesses. The next level is that we started thinking about walls, which were a double line and flat symbols.
“Now with Revit, when you draw a wall, the system knows it’s made out of masonry and knows a lot about its function and structural and thermal performance. The software knows how much it costs and how much of it there is. All that is now embedded in the model.”
And so, with the move to 3D, Autodesk introduced new code-streams like Inventor, and started to broaden outside of the AutoCAD verticals but still maintained strong links between the 2D drafting products and their new 3D modelling tools. Autodesk also started to acquire 3D companies like Alias, bringing yet more technology and customers to the company.
Autodesk’s acquisition of Revit in 2002 for $133 million was a significant moment for the company. It was a new code-base for an AEC solution and offered a parametric 3D modelling system, similar to Inventor. In many respects, Revit finally delivered on the promise of early UK-developed software like Sonata.
“Neither the market nor the products were ready at that time”, commented Bass,”Those systems made many of the promises back then that products like Revit actually deliver on. It was partly because of the products, but mainly because of the computing infrastructure of the time. You have to remember that the machines back then had less power than an iPhone and they cost $100,000. 2D CAD was appropriate for the technology we had and the way we did business, which was through 2D drawings. Now the operating model for design, is through models and that’s across all our disciplines.
“The success of Revit speaks for itself. At the time, people deemed it a risky investment but I thought it was game-changing technology. If you look back seven or eight years ago, the product was in its infancy. The things Revit was capable of doing then would satisfy 1% of the AEC market but with years of development now it’s hugely successful.”
The success of Revit speaks for itself
Bass also pointed out that most of the mechanical parts that were modelled had fewer parts but were geometrically complex. While in AEC, buildings would traditionally be made of millions of parts but all of relatively simple geometry. Today, the lines are being blurred, in both design and fabrication methodology. He added, “What’s interesting is the founders of Revit had a background in mechanical design and so they understood the technology deeply and borrowed or invented a parametric system for the way architects work.”
So with a shrinking economy and tough times for architectural firms, making a transition to using Revit over AutoCAD, making an investment, would appear a more difficult decision. However, Bass had his one take, “The idea that you are not going to do design in 3D is a certainty for putting yourself out of business. There are many ways to go about figuring out how to incorporate model-based design into the workflow but many companies are doing that. Tough times require companies to be more competitive, I think staying uncompetitive is not a formula for success.”
In the past, Autodesk had been a fairly secretive company keeping all new products under wraps. However, now with Autodesk Labs, new technology and demonstrations are available to anyone. Bass explained the philosophy behind the site, “[Autodesk Labs] gives us a quick way to get feedback from customers. Even if the product is good or bad, just being able to get in touch with the early adopters, willing to put up with immature software. From labs, products like SketchBook Mobile is flying ‘off the shelves’, Inventor Fusion has been a big success and so too has Project Dragonfly.” Although there have been a few misses along the way, with Project Freewheel, a web-based DWF viewing tool, hasn’t really gained much traction. Bass consoled himself, “ I don’t really know why but that’s why we have the site!”
Autodesk is currently looking to trademark the DWG format. This would enable the company to stop competitors using the term or writing interpretations of the file format. There are a number of legal steps to go through and it appears to be a multi-year process. It would be akin to Microsoft trademarking the .DOC format and has some industry watchers concerned.
“The issue isn’t really one about trademarks,” said Bass, “ The real issue is about the integrity of the file format and whether there is interoperability. If we started calling our drawing files .DOC, that would be problematic for people and not a good thing. We have been pushing for the open sharing of file formats between companies and we have done deals with PTC and Bentley who are competitors of ours, because I don’t think formats are a point at which we should compete. We have offered the same to every other CAD vendor and I’m happy to do that because I just don’t think it’s a point of differentiation and it’s an annoyance to customers. We need to compete on another axis. The important thing is to be able to read and write the file formats that your customers want you to.”
There have been many rumours that Autodesk is about to go into ‘retail’. At the moment the products are mainly at the high-end professional range of design and even AutoCAD LT, which started life as a few hundred Euros is now well over 1,000. I asked Bass what Autodesk’s pretensions were to retail. He replied, “I’m not sure the right question is about retail channels. Nowadays, the challenge is to find the appropriate channel for each audience. For example, Project Dragonfly and Showroom are web services brought directly to the consumer and more likely than not, these people are not design professionals - but no reason they can’t be.”
It seems Autodesk is looking to develop web-based products like Project Dragonfly (which is available free on the Labs website) and similar for non-professionals that require some kind of visualisation or house planning utilities. There are still continuing rumours of a light version of AutoCAD LT, however ironic that sounds- a light version of a light version.
The web is here to stay and I find it frustrating that many of Autodesk’s applications can’t be bought and downloaded online. A handful are available, but these days when multi-gigabyte games can be bought and downloaded, design applications need to catch-up. The recent introduction of Project Twitch provided a glimpse that Autodesk is looking beyond even the need for a download. Bass said, “We want to be able to deliver all of our products through electronic download and are working towards doing that. We also are piloting a version of SaaS (Software as a Service), with Project Twitch. Based on video compression technology, it potentially can be used to deliver high quality experiences on even less expensive client side hardware than a traditional PC.
Right now, its utility is limited by its proximity to the data centre but the current limited pilot should suffice to get good user feedback. We will also be releasing more native and hybrid SaaS applications to take full advantage of the strengths of each platform. And finally, we will continue to expand to other platforms such as the iPhone and the Mac.”
Any move by Autodesk to provide products and services online will undoubtedly raise eyebrows within its extensive distribution and dealer channels. Bass took issue, “People seem to have trouble reconciling the role of a channel with SaaS but you need to look no further than salesforce.com to realise a channel is a means to reach customers regardless of whether the offering comes in a box or online.”
Reflecting on making the transition from DOS to Windows, Bass identified the stakes are high, with the industry fast approaching a major transition point, using the Internet as a platform. “There’s no doubt that the next transition is from Windows to a cloud-based computing environment. And, the two things that makes market leaders vulnerable are inflection points in the technology and changes to the business model.” Bass explained, adding “ If you want to attack an industry-leading company when the business model and technology are stable, it’s a very difficult thing. A frontal attack offering the same, a little bit cheaper, a little bit faster, most often has proven to be a distinct failure.”
Bass has been a key instigator in the ‘greening’ of Autodesk, not just in the capabilities of its products but also in how the company runs; from water and electricity usage to the overall carbon footprint of the firm when doing its business. With so many engineers and architects using its software tools to design ‘stuff’, Autodesk is looking to provide those designers with tools to limit their environmental impact in the process.
I believe it is critical that architects, engineers and designers to make sustainability a key trait of their work
Bass explained. “I believe it is critical that architects, engineers and designers to make sustainability a key trait of their work. Right now, most products are much more damaging to the environment than they need to be given today’s technology. Consistent with our mission to help people design better products and projects, we will provide software that helps people better understand the environmental implications of their design decisions. Today software is tremendously under-utilised whether an architect is designing a building, an engineer is designing a power plant or a designer is making a new consumer product. Unfortunately, it is not clear that even so-called green projects are substantially more energy efficient than non-green ones. Given the advances in material sciences, the science of systems engineering and the power of software to simulate different designs, this is a tragedy. We will continue to try and build more sophisticated but simpler and better integrated tools so people can make the best possible choices.”
I’ve often wondered that as Autodesk has something like $1 billion in the bank, if it would ever be interested in acquiring a business in a totally different market. A market that was less mature and potentially more profitable?
Bass quickly hit that one out of the ball park, stating, “I have been steadfast that our mission is to help people design better projects and products. Tangentially, we may touch other areas but I don’t see those areas as key to our future.”
I guess if I were one for analogies, the first two decades for Autodesk were like a tree growing a strong trunk and root system, so eventually it could support the great canopy of products that are now emerging from the company. But I feel it also reflects the two characters of Bartz and Bass, with Bartz building a very strong business machine, forcing the company to form branches to better tackle the various industry markets, while Bass has built on that base and further refined and increased those original branches, nurturing new green foliage while seeding markets as the company evolves.
The company is not as stock market driven as it once was and there is certainly a higher concentration of effort in product
development and technology. Autodesk appears aspirational; it has never been in so many new markets, or offered higher degrees of functionality, and is now writing for a wide range of platforms like the iPod, as well as developing new web-based services.
The creation of Autodesk Labs to ensure the first release of software isn’t incomplete or a disappointment but also invites customers in to share their opinions is a brilliant success. Although that’s not to say the business model and yearly upgrade cycle doesn’t still cause some resentment, especially in older products which may not have any obvious development velocity, or customer who’s project cycles have multiple years.
As we were going to press, within the same week, Autodesk released a brand new multi-touch sketch tool for the Apple iPhone (SketchBook Mobile), followed the next day with the announcement of Project Twitch- a trial to run web-hosted AutoCAD, Revit and Inventor sessions. As a lover of technology, things really don’t get much better than this and I’m looking forward to seeing where Autodesk thinks design technology will be going next.