The 1% Horizon: Why Canva Is Betting on Goal-Oriented AI

The "SaaSpocalypse" theory suggests that general-purpose AI models will turn software into a commodity and make specialized platforms obsolete. In a conversation with Bob Safian, Canva CPO Cameron Adams explains a different reality: the most resilient companies are not fighting this change. Instead, they are using it to get closer to what their users actually want to achieve. By shifting from a design-first to a goal-first approach, Canva is not just surviving the AI transition. They are using the friction of rapid iteration to build a deeper, more specialized competitive advantage. For leaders, the edge comes from creating systems where AI amplifies human intent rather than replacing the creative process. This requires moving away from business as usual toward structured experimentation that values outcomes over output.

The Hidden Cost of Business as Usual

Most organizations treat AI as an add-on, hoping to gain efficiency without changing their core workflows. Adams argues this is a mistake. The tendency to rely on familiar tools creates a mental ceiling that prevents long-term improvement. Canva responded to this inertia with "AI Discovery Week," a system-wide pause where employees were told to stop their standard work and experiment with AI tools.

The results were significant: nearly half of the 500 plus projects started that week moved into production. By creating a temporary, safe space for experimentation, Canva bypassed the natural resistance to change. This reveals a clear dynamic: if you do not explicitly create time for exploration, your team will naturally stick to the status quo, which insulates the organization from the innovation needed to survive a platform shift.

"I am sure we are all terribly busy and trying to get 101 things done every single week and you naturally fall into the patterns that you would normally do things in. You have got your tried and trusted tools, you know how they work, you will always reach for them. So you are not looking for that medium term and long term improvement that you can get through some of these tools."

-- Cameron Adams

Why the Goal-Oriented Pivot Creates a Moat

The standard approach in software is to provide users with a tool and let them figure out how to use it. Canva’s shift toward goal-oriented AI, where a user asks "how do I get more customers?" rather than "how do I design a flyer?", changes the entire feedback loop. When a platform moves closer to the user’s ultimate goal, it becomes harder to replace. A design tool is a commodity; a business-growth partner is an asset.

This pivot requires shifting the focus from image generation to design intent. By building design principles and user data into a foundational design model, Canva creates a specialized layer that general-purpose LLMs cannot easily replicate. This is where the competitive advantage grows: the system learns what good design looks like for specific user goals, creating a cycle of better results that keeps users engaged.

"The real power that we have seen with AI is when you can amplify ideas and when you can take what AI gives you and build on top of it, put your own stamp on it, make it authentic and make it connect with the real people that you want to connect with."

-- Cameron Adams

The Architecture of Human-Centered AI

A common concern regarding AI is the fear of removing human judgment. Adams addresses this by ensuring that every AI-generated output in Canva is editable. This is a deliberate choice: by keeping the human in the driver's seat, the system ensures that AI acts as a collaborator rather than a replacement.

The implication is that the most successful AI products will be those that treat the AI as a conversational agent, not a black-box generator. Designing a conversational interface is a challenge, but it allows for a more empathetic interaction where the AI can explain its choices, such as font or color, which educates the user and improves their taste over time. This creates a lasting advantage: the product does not just deliver a result; it improves the user's ability to create.

Key Action Items

• Implement Discovery Sprints: Dedicate one week per quarter where teams are forbidden from business as usual to experiment with AI tools relevant to their specific departmental problems. (Immediate investment)
• Audit for Goal Proximity: Map your product features to your users' ultimate business outcomes. If you are only solving for the how and not the why, you are vulnerable to commoditization. (12-18 month horizon)
• Decentralize Tool Selection: Allow teams to choose their own AI tools based on their specific needs rather than mandating a top-down stack. This fosters an experimental mindset. (Immediate action)
• Architect for Editability: Ensure that all AI-generated content is fully editable and collaborative. If your AI output is a dead end, you are failing to amplify human creativity. (Ongoing product strategy)
• Build a Design Model Equivalent: Identify the core principles of your industry that are currently tacit knowledge and work to bake them into a foundational model for your internal workflows. (12-18 month horizon)
• Prioritize Human-in-the-Loop Interfaces: When designing AI features, focus on interfaces that allow users to steer the model rather than just prompting it. This creates a more authentic connection to the output. (Next 6 months)