In short, it’s CAD for the way we design, manufacture, and work today: distributed, remote, collaborative, and rapidly iterative. … Imagine the productivity gains if CAD software embraced concepts like document sharing, collaboration, cloud storage, and being mobile-native.
Ben Thompson unpacked the announcement via disruption theory (subscription required), observing that Onshape doesn’t look disruptive at all:
- Onshape is not really that much cheaper than incumbent products ($1,200 annually for the professional version and “Contact Us” — i.e. a lot more — for the enterprise version, as compared to $1,680 annually for AutoCAD)
- Onshape is targeting established enterprises that use AutoCAD1
- It remains to be seen if AutoCAD’s most profitable customers want Onshape
I agree and don’t think Onshape will succeed as a disruptive product.
Users aren’t buyers
I first heard about Onshape from mechanical engineering colleagues. The official demos show a transformatively superior collaboration and interaction paradigm, but CAD is overwhelmingly enterprise software. As Ben has pointed out repeatedly on exponent, enterprise software competes with feature checklists, cost, and other spreadsheetable metrics — not user experience.
A couple of early adopters may get excited, but even at a small company switching is expensive. CAD is often tied up with other solutions like configuration and data management or Enterprise Resource Planning (ERP). It’s hard to justify experimentation with new tools if they are going to break existing workflows. Furthermore, CAD software is notoriously awful at common file formats. Each program is proprietary, and exchanging data between programs in pseudo-universal formats like STP is labor intensive, error prone, and typically results in data loss.2 Given the number of file formats Onshape can import and export its experienced team clearly understands this. The quality of these tools will be fundamental to widespread adoption.
Cloud-based tools are a non-starter in the aerospace and defense industry because of International Traffic in Arms Regulation (ITAR). Non-US persons cannot have access to export-controlled data (like CAD models and drawings). This means it can’t be stored in cloud services potentially staffed by foreign persons, or that locate servers or route traffic outside the US. In response, these industries self-host or use specialized, ITAR-compliant backends like Amazon Gov Cloud. The worldwide design industry is large enough to support OnShape without aerospace and defense, but the industry solves some of the most challenging and ambitious CAD design problems. Ignoring this industry would be like Apple writing off artists, designers, or developers.
If OnShape succeeds (and I hope it does), it won’t be as a disruptive entrant, with an inferior but cheaper product that eventually becomes good enough. If it looks inferior on paper it will struggle to get traction in the enterprise, especially if switching costs are high or if adoption is prohibited by federal law. The path to success is as a fundamentally superior product with better version control, improved collaboration, and support for new input devices (e.g. iPad, touch, Pencil). That is, it will win the way Apple wins: superior user experience and just enough enterprise support that the IT department and the CTO will be forced to change by the lamentations of their employees and the promise of increased efficiency.
Some relevant technical pedantry: OnShape’s competition is the 3D modeling engineering market (e.g. SolidWorks, ProEngineer, and Catia), not the 2D Autocad architecture market. The point remains valid. ↩
Even switching between major versions of ProE can be a major initiative. For an organization with libraries of models build in heritage tools, moving to a new system can seem overwhelming. ↩