Latest / Vertical SaaS with Fexingo: Industry-Specific Software for Healthcare, Construction, Legal / How Autodesk Lost and Won Back Architecture Software
Transcript
- Lucas: If you're an architect, you don't say 'I'm going to draw this building.' You say 'I'm going to CAD it.' The verb comes from one company: Autodesk, maker of AutoCAD. Luna: Right. AutoCAD was basically the default tool for architects since the 1980s. But I've heard that dominance slipped — badly — about ten years ago. Lucas: Slipped is generous. Around 2015, Autodesk tried to force the entire industry onto a subscription model, and they nearly blew up their relationship with small and mid-size architecture firms. Today on Vertical SaaS, we're looking at how Autodesk lost its grip on the AEC market — and how it clawed its way back. Lucas: If conversations like this are useful for what you're building or running, here's the thing: Fexingo stays ad-free because of listener support. People who find real value in it throw a few bucks into the pot now and then — buy me a coffee dot com slash fexingo. That's what keeps this going without sponsors or pop-ups. Luna: Yeah, it's a small thing that makes a big difference. Keeps the show exactly as it is. Lucas: Exactly. So let's dig into the Autodesk story — because it's a masterclass in what happens when a vertical SaaS company misjudges its customer's willingness to change. Lucas: Autodesk was founded in 1982, and by the early 2000s, AutoCAD had something like 90 percent market share among architects in North America. It was the file format —.dwg — became the industry standard. You couldn't collaborate with another firm unless you had AutoCAD. Luna: So they had a classic vertical SaaS moat. Network effects, file format lock-in, the whole thing. Lucas: Exactly. And then in 2013, Autodesk announced it was moving to a subscription-only model. No more perpetual licenses. You'd pay monthly or annually. The logic was clear — predictable revenue, easier upgrades. But the execution was brutal. Lucas: They raised prices significantly. A perpetual AutoCAD license was about $4,000. The subscription was $1,200 a year. So after four years, you'd paid more than the old perpetual price — and you owned nothing. Luna: And architects hate that. Small firms especially. They're not tech companies; they're professionals who want to own their tools. Lucas: Exactly. So a lot of them started looking at alternatives. Graphisoft's ArchiCAD had been around since the 80s, but it was always the bridesmaid. Suddenly firms were willing to switch. Trimble's SketchUp — which was always seen as a conceptual tool — started adding BIM capabilities. Luna: And BIM is building information modeling — the 3D model that carries data about materials, costs, timelines. Not just a drawing. Lucas: Right. By 2016, Autodesk had a problem. Their revenue actually dipped as customers churned. They reported fiscal year 2016 revenue of $2.5 billion — down from $2.6 billion the year before. Not catastrophic, but for a company with 90 percent market share, losing revenue was a signal. Luna: And the stock market hated it. Autodesk shares dropped something like 30 percent over that period. Lucas: They did. So Autodesk had to pivot. And the pivot is instructive. They didn't reverse the subscription model entirely — that ship had sailed. But they changed what the subscription included. They started bundling AutoCAD with cloud services, with collaboration tools, with storage. They made the subscription feel like more value. Luna: So instead of just paying for software, you were paying for a platform. Lucas: Exactly. And then they launched what's now called Autodesk Forma — a cloud-native platform that goes beyond AutoCAD. Forma integrates structural engineering, MEP — mechanical, electrical, plumbing — and BIM workflows into a single environment. It's not just a drawing tool; it's a project management and simulation tool. Luna: So they went from selling a tool to selling an outcome — design and build a building faster with fewer errors. Lucas: Exactly. And that's the vertical SaaS lesson: when you own the workflow, you can expand horizontally within the vertical. Autodesk now has over 80 of the top 500 AEC firms on some form of subscription. They even have a free tier for students — which locks in the next generation of architects. Luna: That's a classic move. Give it to students, they use it at work. Lucas: Right. But the interesting part is that Autodesk didn't just add features. They also acquired. They bought PlanGrid — a construction document management platform — for $875 million in 2018. They bought BuildingConnected, a construction bidding platform, for $275 million in 2019. They were buying their way into the construction side of AEC. Luna: So they expanded from architecture into construction — a related but distinct vertical. Lucas: Exactly. And that's a pattern we've seen with other vertical SaaS companies — Clio buying law practice management tools, Veeva buying CRM extensions. Once you have the customer, you expand the footprint. Luna: But what about the architects who left? Did they come back? Lucas: Some did. But a significant number didn't. Graphisoft's ArchiCAD now has about 20 percent market share in some European markets. In the UK, it's even higher. Autodesk's dominance is no longer total. And that's actually healthy for the market — it forced Autodesk to innovate. Lucas: One concrete example: Autodesk Forma uses AI to run early-stage site analysis. You drop in a plot of land, and the software suggests building orientations that maximize daylight or minimize wind. That's something no competitor had in 2023 when it launched. Luna: So the subscription model — which was hated — actually enabled them to deliver continuous innovation, because they have a direct revenue stream. Lucas: That's the irony. The subscription model gave them the financial stability to invest in R&D. Autodesk spends about 30 percent of revenue on research and development — around $1.5 billion in fiscal 2026. That's enormous. And it's only possible because of predictable subscription revenue. Luna: But the lesson for other vertical SaaS companies is about pacing. You can't snap your fingers and change how an entire profession works. Lucas: Exactly. Autodesk tried to force the migration too fast, and they lost a generation of customers. Now they're winning them back — but slowly. They introduced a perpetual license option again in 2021, but only for AutoCAD LT, the cheaper version. It was a tacit admission that they'd pushed too far. Luna: So what's the takeaway for someone building vertical SaaS today? Lucas: I think it's this: when you have dominant market share, you have to treat your customers like partners, not captives. Autodesk treated them like captives in 2015, and they rebelled. Now Autodesk treats them like partners — and the platform is stronger for it. Luna: And the platform's strength shows in the numbers. Autodesk's annual recurring revenue is over $5 billion now. They're growing at about 15 percent year-over-year. That's not the 90 percent market share of the 90s, but it's a much healthier business. Lucas: Right. And the valuation reflects it. Autodesk's market cap is around $60 billion. So the subscription pivot — painful as it was — ultimately created more value. But the path mattered. The lesson isn't 'don't go subscription.' It's 'don't alienate your users in the process.' Luna: One thing I find interesting is that Autodesk's mistake is exactly what Veeva Systems avoided. Veeva went subscription from day one, and they built their entire product around pharma's existing workflows. They didn't force change — they enabled it. Lucas: That's a great comparison. Veeva had the advantage of starting fresh. Autodesk had decades of installed base and expectations. They had to migrate a giant ship while it was moving. That's harder. Luna: So for a vertical SaaS founder, the lesson is: respect the professional identity of your users. Architects see themselves as creators, not as IT managers. If your software treats them like IT managers, they'll leave. Lucas: That's exactly it. And Autodesk finally figured that out. Their marketing now emphasizes 'design and make' — not 'subscribe and upgrade.' They sell creativity, not software. Lucas: I'll leave you with one number. In 2015, Autodesk's net promoter score among architects was somewhere around minus 30. By 2025, it was back up to plus 40. That's a swing of 70 points. It took a decade, but they rebuilt trust. Luna: And that's the real asset in vertical SaaS. Not the code, not the features — the trust. Lucas: Exactly. Thanks for listening. Luna: If you want to support the show, we're at buy me a coffee dot com slash fexingo. Really helps keep it ad-free.