The Exodus Curve

The concept of Product-Market Fit (PMF) collapse has gained renewed attention with the rise of large language models (LLMs), as highlighted in a recent Reforge article. The article argues we’re witnessing unprecedented market disruption, in this post, I propose we’re experiencing an acceleration of a familiar pattern rather than a fundamentally new phenomenon.

Adoption Curves

Before starting, it is worth revisiting the classic technology adoption curve (sometimes called diffusion of innovation). The curve is defined by a predictable sequence of people trying a new product or service:

  1. Innovators: The risk-takers who test and validate new ideas.
  2. Early Adopters: Visionaries who bring early momentum.
  3. Early Majority: The pragmatic segment that brings scale (PMF happens here).
  4. Late Majority & Laggards: The last to adopt, often driven by necessity.

Startups internalize this strategy by relentlessly seeking out the innovators who are happy to deal with some janky attributes of a solution because they can see the step function change that, if functional, this radical new approach may bring. Innovators can see through the flaws and recognize the potential of something not fully built.

The Exodus Curve

It stands to reason that the same early adopters coming to our new solution are leaving behind some incumbent substitute. From the lens of the incumbent they observe an exodus curve.

As far as I’m aware no company is categorizing users in terms of their innovation curve behavior. Once a user enters as product they become a user, maybe with a specific role, job title, or JTBD (job-to-be-done) but we do not label their psychographic behavior.

If we invert the adoption curve to represent an exodus curve we’d see the same pattern. Slowly, slowly, then all at once. The same way a new start has a handful of early adopters until they unlock the early majority and see hockey stick growth.

The exodus curve

Historical Patterns of PMF Collapse

“History doesn’t repeat itself but it often rhymes,” – Mark Twain

While AI and LLMs may accelerate the pace, PMF collapse isn’t new. Consider these historical examples:

  • DVD to streaming (Netflix’s strategic pivot)
  • Directory enquiry phone lines to internet search
  • Kodak’s film business to digital photography

The Reforge article highlights how products built around contributor content, e.g. Stack Overflow and Chegg, risk collapse when contributors stop posting. The first nodes in their growth system is the point of failure.

The challenge for incumbents isn’t just identifying the threat – it’s their ability to pivot given the massive resources and processes built around optimizing their existing systems and feedback loops.

Figma: A Case Study in Potential PMF Collapse

Figma, despite its current dominance in the design tooling space, shows early warning signs of PMF collapse. The threat comes from both ends of the design spectrum, here is my new workflow:

  1. Early ideation and exploration is done either on paper or scrappy sketching tools such as excalidraw where speed and divergence of ideas is required.
  2. The emergence of the LLM-powered designer using Cursor and Windsurf to make good enough prototypes is eliminating the need for frame-to-frame simulations in Figma (which still doesn’t allow text input 😅).

These shifts are particularly visible among early adopters on tech Twitter – the same community that championed Figma’s rise.

From my view, it appears Figma’s current business strategy is centred around:

  1. Design teams building a large design system in the tool as a store of value making switching painful
  2. Designers bring engineers into the tool to collaborate with design teams in Figma
  3. Bringing marketing/ non-product people into the tool to brainstorm (Figjam) and create decks (Slides)

The core loop is under threat as innovator designers move out and collaborate with engineers in new mediums. Rather than engineers joining designers in Figma, designers can now join engineers in code.

While Figma might not notice significant revenue impact from this group yet, could innovator exodus be the pre-cursor to PMF collapse?

One thing is clear, PMF isn’t a permanent achievement but a dynamic state requiring constant monitoring and evolution. In an era of accelerated disruption, the ability to detect and respond to early warning signs of PMF collapse may be the difference between thriving and becoming the next business school case study in disruption.