The Business Model Canvas is a classic for good reason. Nine boxes force you to look at your business as a whole: value proposition, customer segments, channels, customer relationships, revenue streams, key resources, key activities, partners, cost structure. No hiding behind one well-optimized corner.
But filling out a BMC in 2026 isn't the same thing as it was in 2010. Three things change — and understanding them is what separates good strategy from a time-traveling artifact a 2010s consultant would have produced.
What changed
1. Key activities are partly AI's job now
The "key activities" box used to list what your company does daily: development, customer support, content production, marketing. Each one required people.
Today many of these activities are partly or fully run by AI:
- Content production: AI writes the first draft, humans edit and verify
- Customer support: First-line is a chatbot, escalation goes to humans
- Research and analysis: AI fetches and aggregates, humans interpret and decide
In a BMC this doesn't mean fewer activities. It means each one needs a clear split: which part is AI, which part is human. This is also usually where hidden scalability hides.
2. Key resources aren't just technology and brand anymore
Classic "key resources" included machines, contracts, team, IP, brand. AI adds to the list:
- Prompts and methods: how you use AI. Your "AI workflow" can be as critical as your codebase.
- Data: what proprietary data you have, and how AI can be tuned on it
- AI relationships: which AI providers (Anthropic, OpenAI, Google) — multi-vendor or lock-in?
Ask yourself: if a competitor got the same AI model as you tomorrow, what would set you apart? The answer should show up in your BMC key resources.
3. Value proposition is not "AI-powered"
A common mistake: build a product, add "AI-powered" prefix, call that the value proposition. This doesn't work anymore. In 2026, AI is infrastructure, not a differentiator.
Your value proposition is still what it always was: which customer problem you solve and why your way is better than the alternatives. AI changes how you solve, not what.
A good test: if you removed the word "AI" from your product description, would the customer still understand the value? If not → the proposition is too loose, AI is too big a part of your identity, and every new AI model generation will rattle you.
A practical filling order
I recommend this order especially when filling out a BMC in the AI era:
- Customer segments first — who you're building for
- Value proposition — what problem you solve, in plain language, without the word AI
- Key activities — what has to happen for the value to be delivered. Mark each: human / AI / hybrid
- Key resources — what you need for those activities. Don't forget AI resources (prompts, methods, data)
- Channels and customer relationships — how you reach and keep them
- Revenue streams — how you make money
- Partners — who you do key activities with
- Cost structure — what running the activities costs
This order forces you to solve customer and value before diving into AI specifics. You avoid the classic trap of building an AI solution that has no customer.
How Innovaidor helps you fill out a BMC
In Innovaidor, BMC is one of the methods you can activate during a chat. In practice:
- Write about your idea freely — identify the problem you're solving
- Activate the Business Model Canvas method
- The AI asks one box at a time but challenges your assumptions — it doesn't flatter, it acts as a critical sparring partner
- The output is a document (markdown) that travels with you to Lovable, Claude Code, or any building tool
The point isn't to fill boxes fast. The point is to find the spots where your thinking doesn't survive a reality check — and fix them before you invest time.
Closing
The Business Model Canvas isn't outdated. It's more useful than ever, because AI-era uncertainty makes the fundamentals more important. You don't need a 50-page business plan — you need one A4 where every box survives a hard question.
Start there.