The story around SaaS has turned severe. Investors see AI agents writing code, startups ship internal tools in days, and every executive asks why a company should keep paying for another seat based subscription. The old software bargain feels exposed. If a model can draft text, analyze data, generate interfaces, and automate support, many traditional apps start to look like expensive wrappers around workflows.
That fear is real, but it points to a bigger opportunity. The Figma view is useful because Figma has always sold more than a design canvas. It sells shared context. Designers, product managers, engineers, marketers, and founders gather around the same object, make decisions visually, and keep the work moving. AI makes that shared object more valuable, because more people can create drafts, prototypes, screens, diagrams, and specifications before a specialist polishes them.
This is why the death of SaaS is the wrong lens. The weak layer is routine interface work. The stronger layer is the system that understands a team, preserves its decisions, manages permissions, connects with existing tools, and turns raw output into something editable and trusted. AI can make small tools cheaper, while making serious platforms more central to the way work happens.
Figma is a useful case because its product strategy keeps moving toward the moment where an idea becomes a shared artifact. A prompt can generate a screen. A rough sketch can become a prototype. Marketing teams can create assets. Product teams can test flows. Engineers can inspect logic and implementation details. The value comes from the continuity between creation, feedback, revision, and delivery. In that loop, SaaS becomes the place where human judgment and model output meet.
The first signal is the spread of creation. In the earlier SaaS era, software often mapped a fixed department. Sales teams used one system, finance teams used another, design teams used another. AI changes the shape of participation. A founder can make a prototype before hiring a full product team. A support lead can draft a knowledge base flow. A researcher can turn a chart idea into a publishable figure. This makes specialized tools easier to start using, because the first draft no longer requires a long chain of handoffs.
The second signal is context. A generic chatbot can produce a useful fragment, but the real work needs brand rules, prior decisions, customer data, approval history, file versions, and team roles. SaaS companies that own this context can make AI output fit the organization. They can remember what a team values. They can keep assets compliant. They can make the next draft smarter because the product already knows the previous thousand decisions.
The third signal is editability. AI output is impressive when it appears quickly. It becomes valuable when a person can revise it without starting over. That is why tools for technical and creative work matter. A research team can ask ChatGPT to outline an experiment, compare reasoning with Gemini, use Miss Formula to turn formulas from images into editable notation, and use Editable Figure to convert AI generated paper figures into editable vector graphics. The winning workflow is the one that lets humans take ownership of the result.
This changes the business model question as well. Seat based pricing will feel strained when AI does more of the clicking. Yet SaaS can price around outcomes, usage, collaboration, governance, or the value of finished assets. A product that saves a company from messy files, broken handoffs, legal risk, and repeated manual cleanup still has pricing power. The value moves from access to acceleration and reliability.
The best SaaS products of the next era will feel like studios, memory systems, and operating layers at the same time. They will give people a fast way to make something, a safe way to refine it, and a reliable way to ship it with others. Figma sees this clearly because its product has always been a room where work becomes visible. In an AI world, that room becomes more important, because creation gets faster and coordination becomes the scarce resource.
So the real lesson is simple. AI will pressure any SaaS product that only stores fields and waits for clicks. It will lift products that own context, collaboration, permissions, memory, governance, and the final editable artifact. SaaS is entering a harsher market, but also a more imaginative one. For companies that can turn AI output into shared, editable, trustworthy work, this may be the best era software has ever had.
Top comments (0)