Entering Tannen’s Magic Store, a New York City landmark of the prestige scene, feels like walking into a 1990s VHS rental store to grab a movie on a Friday night. The man behind the counter sports a half-zipped sweatshirt, a ponytail, and a skull ring. He pulls decks of cards and boxed tricks off the shelves, demoing them for tourists and fellow magicians looking for a new taa daa moment. He shrinks a penny to half its size and illuminates a lightbulb off of a lamp like a cigarette.

“I used to demonstrate a similar version of that trick,” Jon Hirschtick muses as he watches a quarter spin in mid-air to gasps from the store patrons.

Hirschtick may be the founder of two product manufacturing system companies, but he cut his teeth on smoke and mirrors. In the 1980s, he was both a professional magician and original member of the MIT Blackjack team, which inspired the movie 21. We’ll leave you to separate the fact from fiction in that film, but Hirschtick was able to use $1 million of his winnings to start SolidWorks, a software company which became the industry standard for computer-aided design (CAD). Now, when most executives would have their eye on retirement, Hirschtick is anteing up again. Aiming to disrupt a market dominated by his own company, Hirschtick has launched a new CAD startup: Onshape.

Why? The opportunity he has spotted is that CAD tools for designing products haven’t really changed in 25 years, leaving engineers to fight everything from install problems to workflows that not only discourage collaboration, but lock teammates out of the design process.

Hirschtick is betting Onshape can unlock the gridlock – and so is Andreessen Horowitz, which led Onshape’s $105 million series D. That might sound like a lot but it’s a drop in the  $8.7 billion CAD industry ocean that almost no one talks about.

Surrounded by fans of cards, top hats, wands, and the coin trick books he studied as a teenager, Hirschtick discussed why he’s trying to disrupt the same market twice, what it takes to be a great designer, and how to wow your audience every time.

A successful card player might get up from the table after playing a successful hand. After selling SolidWorks to Dassault Systèmes for $318 million, why get back in the game with Onshape? 

All engineers, me included, try to see a better world through some sort of product.  Once seeing that vision of the future, engineers are very driven to realize it. Our job building great CAD was far from over. I felt an obligation; nobody else has shipped a true modern CAD system besides Onshape.

What exactly are you making with Onshape?

CAD are computer software tools that are used by engineers to design things. We’re a meta-designer. Engineers have always had tools. In the age of the pyramids it might have been papyrus, then paper, quill, ink. 50 years ago, it was pencil and paper. Today, most engineers use some form of CAD software. If you’re manufacturing a new product, you build it twice. First you build it in the computer in CAD. You make sure it’s right; the pieces fit together. And then you build it in the real world. A CAD system is like a script or a rehearsal.

Why market need does Onshape address that SolidWorks doesn’t?

I founded SolidWorks 25 years ago with five other people in my home. There’s millions of users who use it. Around five years ago, I’d visit designers and watch people use the system and I could see they had problems. They have problems just installing the stuff. It only runs on Windows. The next set of problems is that the data is stored in files. I don’t mean one file. I mean one file per part. If you’re making a snowmobile with 3,000 pieces, that’s 3,000 files. Everyone needs a copy of the 3,000 files to look at the design. With design the whole goal is speed and creativity. We want to make changes, iterate, and find the best answer.  But the tools say, “Wait a minute, before you make a change, do you have to latest version of, for example, File 1920?” That’s a good way to slow the team down. Things crash, people lose work.

So, those things were a top priority to fix in the launch of Onshape?

We felt we could design something where others could design their stuff faster, have better ideas, and honestly, have more fun. No one ever had any fun typing license codes and overwriting files. We borrowed from Google docs, who inspired us with real time collaboration. You go in and two people are in the document. Now, do that with the 3,000 parts. Those 3,000 parts are now all in one place in the cloud. We can work on it at the same time. We don't have to lock anything and I don’t have to ask, “Where’s the latest version?” If I change the shape of the front fender of the snowmobile and you’re 1,000 miles away, you see if instantly. We allow concepts of branching and merging. So, I can try five different ideas and I don’t have to worry that I’m overwriting work.

Looking at how the industry has changed since you founded SolidWorks, is an engineer or a designer’s job easier or harder now?

Harder. You have to master many more technologies than you used to. Engineering is a broader subject with fewer clear lines of demarcation. Same with design. As a designer, you still have to know about print, because there’s a lot of a lot of print in the world. But you also have to know all about computing platforms, HTML, mobile devices. It’s an incredible palette of technologies you have to learn.

Where do you think the future of design is headed?

I think more of the world’s GDP is generated and differentiated by design than ever before. I’m not just talking about competitive differentiation. I’m talking about how the number of products is exploding. If you go to the store to buy laundry detergent, when I was a kid there was a ‘big box’ and a ‘small box’. Today, there’s 41 products from Tide: plastic bottles, the Tide Kick you throw in, a special spray. A category that had a few products, now has an enormous amount. There’s so much more choice. The SKU explosion is crazy. You used to go to the store to buy a stapler and it was a bent piece of metal.  Today, there’s three different cool designs in 18 colors. So, I believe the amount of design as a share of GDP is growing.

After 37 years in the CAD business, what advice do you have for designers?

A designer is never executing the status quo. Good designers are always moving forward. Many people have great jobs, but they don’t change the way people do things; they don’t envision a different world. Like, if you run the train system in New York, your job is to keep those trains moving from the 34th Street station to the 14th Street station. Your job is not to move the stations around. Zero designers are just operating in the existing world. You have to see something that doesn’t exist. You have to engage in creative hallucination. Visions and hallucinations look the same until you try to build them.

Was your background performing magic tricks useful to you as an entrepreneur?

Magic is where I learned to demo things. When you start a company, you have to demonstrate the product. The demos need to be interesting, clear, exciting, and you want them to work. When I ended up building design projects at MIT, my friends would make elaborate designs that they could never get to function. They didn’t really appreciate what, to me, seemed obvious: the premium on shipping something that worked as opposed to fooling around with big ideas.

With demos, there are a few things to keep in mind. Practice it so it works. Most people don’t practice enough. Part of it is so it’s really reliable, and part is so that you have plenty of mental capacity left to talk. If you’re too dominated by remembering which button to press, then you won’t be able to think about the ‘patter’ of presentation. Practice a lot more than you think and practice from a cold start.

Anything else?

Have a backup plan. A trick I learned from magic is: don’t tell people in advance what you’re going to do. For instance, I can show you a trick with a deck of cards and I might ask you to do something which requires you to make a choice. If you pick one choice, I’m going to do a mind-blowing effect. If you pick the other one, I’ll still do a good magic trick. But you don’t know which I’m going to do. Similarly, I carry a video when I demo Onshape. If my browser crashes, I can bail to that video. You don’t know something’s gone wrong because I never said whether I was going to show you a video or not.