AI Canvas1. Desirable
Is it worth it?
- 1. Beneficiaries
- 2. AI value
- 3. Point of integration
- 4. Human-AI relationship
Learn
Nine blocks to look at the work before the tool
The AI Canvas keeps the structure and the layout of the Business Model Canvas. Each block takes on a new meaning, but the logic stays the same: desirable on the right, feasible on the left, viable at the bottom.
AI Canvas
Beneficiaries
Who benefits from the improvement: the employee, the client, the manager?
AI value
Which pain do we relieve? Which gain do we create?
TasksWhat the role really does.
PainsWhat slows down, repeats, wears out.
GainsWhat we hope to get.
MeaningWhat we love to do and protect.
Point of integration
Where does AI fit into the daily routine: email, business software, Teams, a form?
Human-AI relationship
Automate, augment or assist? Who validates? What stays human?
Target process
Which steps change, before and after?
Data and tools
What data do we need, in what state, accessible to whom?
Allies and dependencies
Who needs to be on board: IT, the vendor, an internal champion, the union?
Costs and risks
Subscriptions, training, compliance, the cost of a mistake?
Measurable gains
Hours recovered, errors avoided, delays shortened. And where does that time go?
- Desirable
- Feasible
- Viable
- Meaning
What we borrow from Osterwalder
The Business Model Canvas, published in 2010 by Alexander Osterwalder and Yves Pigneur, comes down to one idea: put a complex system on a single page, in nine blocks, so everyone around the table sees the same thing.
Their Value Proposition Canvas adds a reflex most AI initiatives lack: start from the need. Not from the solution. Describe the tasks, the pains and the gains before talking about the offer.
Three adjustments for AI
- 1The customer becomes a roleWe don’t analyze the whole organization. We analyze a specific role or process. One canvas per role.
- 2The offer becomes a modeAutomate, augment or assist. Choosing the mode is already half the job.
- 3A place for what stays humanThe blind spot of the original canvas. Yet that’s exactly where team buy-in is won or lost.
Three zones, three questions
Read the canvas like a map. Each zone asks a different question, and you fill them in this order.
2. Feasible
Can we do it?
- 5. Target process
- 6. Data and tools
- 7. Allies and dependencies
3. Viable
Does it hold up?
- 8. Costs and risks
- 9. Measurable gains
The nine blocks
1
Beneficiaries
In the BMC: Customer Segments
The question
Who benefits from the improvement: the employee, the client, the manager?
We don’t analyze “the organization.” It’s too broad to produce anything but generalities. We analyze a specific role: the administrative assistant, the payroll team, the field advisors. One canvas per role or per process. Then name everyone who benefits from the improvement. There are often more of them than you think.
2
AI value
In the BMC: Value Propositions
The question
Which pain do we relieve? Which gain do we create?
The heart of the canvas. Here you describe the role’s tasks, its pains and the hoped-for gains, then a fourth zone we always add: meaning. What the person loves to do. What makes them proud of their work. We name it before we talk about automation.
- TasksWhat the role really does.
- PainsWhat slows down, repeats, wears out.
- GainsWhat we hope to get.
- MeaningWhat we love to do and protect.
3
Point of integration
In the BMC: Channels
The question
Where does AI fit into the daily routine: email, business software, Teams, a form?
The most underrated block. A solution that forces people to open a new tool, log in somewhere else or copy and paste between two windows will die of friction. The best AI is the one you meet where you already work.
4
Human-AI relationship
In the BMC: Customer Relationships
The question
Automate, augment or assist? Who validates? What stays human?
AI doesn’t do just one thing. It can automate (it runs on its own), augment (it prepares, the human decides) or assist (the human does the work, AI helps as needed). Choosing the mode is already half the job. Also write down who validates and what will never go out without a human signature.
5
Target process
In the BMC: Key Activities
The question
Which steps change, before and after?
Describe the process as it’s really done, not as the procedure manual says. Then the after version. The gap between the two shows what people will need to learn, let go of or protect.
6
Data and tools
In the BMC: Key Resources
The question
What data do we need, in what state, accessible to whom?
The uncomfortable question: does the information really exist, or does it live in one person’s head and in three Excel files that don’t talk to each other? This block also forces you to face privacy law head-on, under legislation such as Quebec’s Law 25 or GDPR: as soon as a process touches personal information, you need to know where the data goes, who has access to it and whether a privacy impact assessment is required. Better to find out in a workshop than after signing a subscription.
7
Allies and dependencies
In the BMC: Key Partners
The question
Who needs to be on board: IT, the vendor, an internal champion, the union?
No AI project succeeds alone. Name the people and organizations you depend on: IT, the software vendor, a credible internal champion, the union, the person in charge of protecting personal information. And the ones who can stop everything.
8
Costs and risks
In the BMC: Cost Structure
The question
Subscriptions, training, compliance, the cost of a mistake?
The visible costs (licences, setup, training) and the hidden ones: time spent cleaning data, compliance and, above all, the cost of a mistake. A wrong answer to a citizen and a pricing error in a quote don’t carry the same weight.
9
Measurable gains
In the BMC: Revenue Streams
The question
Hours recovered, errors avoided, delays shortened. And where does that time go?
Write down what you can actually measure: hours, turnaround times, errors. Then ask the question almost nobody asks: where does the time we get back go? Without an answer, it fills up with email and meetings. With an answer, it goes back to what matters.
What are we protecting?
The meaning zone
In the AI value block, we always add a fourth zone to tasks, pains and gains: meaning. What the person loves to do. What makes them proud of their work.
We name it before we talk about automation. Not to look nice. So people know what we’re protecting before we ask them what we’re changing.
The sorting matrix
Before the canvas: pick the right processes
Filling in a canvas takes a good hour of discussion. You don’t do it for every task in the organization. First, sort along two axes: the pain level (repetitive, slow, error-prone or mentally draining) and the human value (judgment, relationships, empathy, meaning).
HighPain level →Low
We automate
The chore
Data entry, report formatting, email sorting, appointment booking. Nobody will cry if AI takes these over. The ideal ground for early quick wins.
We augment
The noble burden
Case notes, inspection reports, credit file analysis. Heavy work, but it takes professional judgment. AI prepares a first draft, the human keeps the pen and the responsibility. This is often where the biggest gains lie dormant.
We wait
The noise
Tasks that are neither painful nor meaningful. You can get to them later. Or never.
We protect
The sanctuary
The heart of the job: welcoming a young person in difficulty, negotiating with a key client, the team meeting where tension gets resolved. AI has no business here, except to free up time so people can devote more of themselves to it.
LowHuman value →High
This matrix does something the canvas alone doesn’t: it makes explicit what we refuse to automate. And that changes the whole conversation with teams.
Ready to try it?
The facilitator guide covers all of this in a printable format, with the timed agenda for a three-hour workshop.