Three AI Pillars

"We approach AI inside the product through three pillars. First of these is that we need to build some of our own AI tech... Second pillar is just finding the world's best AI people to partner with... And for us, the third pillar is our app ecosystem." - Cameron Adams

What It Is

Three AI Pillars is Canva's strategic framework for integrating AI into their product. Rather than trying to do everything themselves or outsourcing everything, they identify three distinct approaches based on where they can create the most value: build proprietary technology where they have unique advantages, partner with best-in-class providers for commoditized capabilities, and enable their ecosystem to drive innovation at the edges.

This framework helps companies avoid the trap of either over-investing in building everything themselves or becoming too dependent on third-party providers.

How It Works

Pillar 1: Build Proprietary AI

Focus on building AI technology where you have:

• Unique data advantages - Data others don't have
• Deep domain expertise - Understanding others lack
• Critical business importance - Core to your product's differentiation

At Canva: Teams build their own AI models around design and images because they have massive design-specific data and it's central to their value proposition.

Pillar 2: Partner with Best-in-Class

Use partnerships for:

• Commoditized capabilities - Things others can do better
• Areas requiring massive investment - Where you can't compete on resources
• Fast-moving technology - Where keeping up internally isn't feasible

At Canva: They partner with OpenAI for LLMs and RunwayML for video generation rather than building their own.

"There's a bunch of stuff that you don't need to internalize in your company. You don't need to create an LLM because it's a commodity thing now. And there's a bunch of providers who can do it way better and have way more resources to do it with than you do."

Pillar 3: Enable Ecosystem Innovation

Leverage your platform to:

• Give developers access to your audience - Create incentives for others to build
• Enable niche use cases - Things you wouldn't prioritize internally
• Let the market innovate - Find solutions you didn't anticipate

At Canva: With 170 million monthly users, developers eagerly build apps integrating music generators, virtual avatars, and other AI features into Canva.

How to Apply It

1. Map your AI opportunities - List all the AI capabilities that could enhance your product

2. Assess each opportunity against three criteria:

   • Do you have unique data or expertise?
   • Is this critical to your core value proposition?
   • Can someone else do this better/faster/cheaper?

3. Assign to pillars:

   • Build if you have unique advantage AND it's critical
   • Partner if it's commoditized OR requires resources you don't have
   • Enable if it's valuable but niche OR you want market innovation

4. Staff accordingly:

   • Hire/build ML teams for Pillar 1
   • Develop partnership capabilities for Pillar 2
   • Build platform/ecosystem team for Pillar 3

5. Maintain coherence - Ensure all three pillars create a unified user experience

When to Use It

Use this framework when:

• Planning AI strategy - Deciding where to invest in AI capabilities
• Resource allocation - Determining build vs. buy vs. enable decisions
• Team structure - Designing AI and ML team organization
• Partnership decisions - Evaluating which AI providers to work with
• Platform strategy - Deciding what to open up to third-party developers

Source

• Guest: Cameron Adams
• Episode: "Inside Canva: Coaches not managers, giving away your Legos, and embracing AI"
• Key Discussion: (00:56:04 - 00:58:49) - Cam explains Canva's three-pillar approach to AI
• YouTube: Watch on YouTube

Related Frameworks

• Build and Buy - Strategic approach to when to build vs. acquire capabilities