AI Redefining Design Economics
How Figma defied Adobe's bundlenomics
AI image engines turn design from a labor business into a software business. Once a model is trained, the cost of serving one more image is tiny, so a company can sell thousands of generations for the price of one human mockup and still keep software like margins. Each prompt, edit, and selection also creates feedback about what users want, which helps improve defaults, styles, and workflow over time.
-
The user workflow is much closer to search than to hiring. Instead of briefing a designer, waiting, and revising, a marketer types a prompt, gets multiple drafts in seconds, picks one, tweaks it, and exports it. That speed is why AI design tools spread first with non designers and high volume creative work.
-
This is the same wedge Canva used against Adobe, just pushed further. Canva won by making visual creation simple enough for PMs, salespeople, and marketers. AI reduces the skill threshold again, from choosing templates and dragging elements to simply describing the output, which broadens the market beyond trained creators.
-
The margin structure comes from compute economics. Current image API pricing shows generated images can cost only cents each at common quality levels, while subscription products bundle many generations into fixed monthly plans. That creates room for 80% plus gross margins once usage is spread across a large paying base, especially for self serve tools like Midjourney.
The next step is that AI stops being a standalone image button and becomes the default creation layer inside broader design suites. Canva is already bundling AI across presentations, video, and marketing assets, and AI native tools like Gamma show the same pattern in slides. The winning products will own the full workflow, from prompt to editing to publishing, not just the model call.