Mike, I found this piece compelling and I agree with the core thesis. The general-purpose infinite canvas is a semantic dead end. A sticky note at a particular set of coordinates means nothing to a machine.
But this problem has already been solved in at least one domain: construction scheduling.
The Last Planner System software category (Touchplan, inTakt, Outbuild, vPlanner) figured this out years ago. They use a constrained canvas. A card is always a task object with defined properties. A line is always a dependency relationship. The ontology is enforced by design.
Result: when the session ends, no human transcription required. The schedule generates automatically because the software knows what everything represents. Some of these tools export directly to Primavera P6 or MS Project with full CPM logic intact.
The difference isn’t canvas vs. database. It’s general-purpose vs. domain-specific.
Miro lets you do anything, so it understands nothing.
Touchplan only lets you do scheduling, so it understands everything you do.
The real question your article raises for me: why hasn’t anyone applied this constrained canvas model to other domains? Strategic planning, product roadmapping, investment decisions. All the things people currently do in Miro with zero semantic output. The template exists. Why isn’t it being replicated?
Hi George, thanks for your comment.. and I typically come up with things that are absolute common sense to me, so I'm not surprised they are being done already. Why isn't it being done in strategic planning? I think it's because the people involved in those functions are not constrained and have focused all of the attention on themselves. It's theater. They'll tell you" they get energy from the process" and my response would be. I don't care if you get energy from it, because you're actually sucking energy out of it. 😄
I predict that things are going to change dramatically over the next few years as clever people start developing agentic solutions that drive directly to whatever specific outcome is being sought. I think the agents will be smart enough to adapt to different domains as well
I manage the template ecosystem at Mural. I really appreciated hearing your point of view—it was thought-provoking and I found it genuinely interesting. As you might imagine, I disagree with several parts of the argument. A few thoughts after watching your video:
• The goal of a digital whiteboard isn’t efficient data entry. It’s to help people who see the world differently externalize what’s in their heads. Once externalized, those ideas become easier to compare, discuss, and align on. Digital whiteboards are simply the latest extension of a tens-of-thousands-year-old human behavior (e.g., the Caves of Lascaux in France or Altamira in Spain).
• Sticky notes are helpful because they force brevity. Unlike a Google Doc, a sticky note nudges you to distill your thought into a single bullet-sized idea that others can quickly parse. That constraint isn’t skeuomorphic—it’s just a device that helps people communicate more empathetically. (Trello cards work the same way; the “sticky note” metaphor isn’t the important part.)
• Your overall premise seems to be that unstructured whiteboard data is inherently problematic, but I think that reflects a misuse of whiteboards:
a) Your described scenario: Five people generate 100 ideas on a whiteboard, and then all 100 need to be moved into a system of record.
b) How whiteboards are meant to work: A whiteboard helps those five people pull 100 ideas out of their heads, discuss them, narrow them down, and choose the 1–3 that deserve to move forward. Only those selected ideas need to go into a system of record. Most ideas are ephemeral—97 out of 100 shouldn’t live as structured data anywhere. The whiteboard is simply a prop that enables low-pressure collaboration, debate, and decision-making. After years of design work, I’ve come to believe almost every idea belongs in the trash; the job is to find the few that don’t.
• It also seems like you’re viewing digital whiteboards primarily as brainstorming tools. Brainstorming is just one of 100+ activities people do in these spaces. For example, the “jobs to be done” ladder you show at 7:10 works extremely well in a digital whiteboard. (In Mural, there’s a template for this called Whys and Hows, used by thousands of people per month.) My guess is that brainstorming accounts for maybe 10–20% of Mural usage.
• The future you describe at 7:45 makes a ton of sense. Facilitators should only need a few minutes to prep for a session like that—certainly not four hours.
• The “pulling data from systems of record” vision is compelling, but today it requires a huge amount of overhead. Someone has to connect those systems, get approval from a cautious IT team, ensure GDPR compliance, and decide what data should or shouldn’t flow in. It’s still a lot of friction. AI will help lower the burden of installing and maintaining integrations, but for now I’d guess there are tens of thousands of people using sticky notes for every one using a data integration. The scale is currently wildly different.
•Syncing back to systems of record is even harder. In a collaborative tool, it’s difficult to convey clearly what’s being synced and who has control. One person may set up the connection and understand it fully; another might join later and inadvertently change something in Product A, not realizing they’re modifying data in Product B for thousands of users. After 20+ years of designing interfaces, I’d say the level of clarity required here is about as hard as interface design gets. Miro is racing toward this vision, and I think it’s making their product more confusing by the day.
Anyway, thanks again for the thought-provoking video. I enjoyed hearing your perspective, and there’s a lot of wisdom in it—even if I quibble with parts of the argument.
"Sticky notes are helpful because they force brevity" You should really read my article. This is one of several challenges that a first principles analysis would tackle. What is the principle behind brevity facilitating decision-making, and what is specific source. I'll go check it out.
My overall premise is that there is a great amount of waste in the "process" of decision-making. Just like there was a great amount of waste in building rockets. Once the status quo was challenged, we suddenly can get things into space for a fraction of the cost that existed just a few years ago. This is why I recommend reading the entire piece, and maybe some other articles on my blog. I use the exact same FP + JTBD approach to everything.
Ryan, thanks for taking the time. I urge you to read the entire article because the video is a barebones summary.
The point is that the goal ( the outcome) of collaboration is a decision, and workshops are a solution (a means to an end). Facilitators (expensive) are the glue that hold workshops together (and deal with the exact same limitations of a digital canvas).
From an innovation perspective (First Principles + JTBD) this is ripe for disruption in my humble opinion. History has shown us that every industry drives to zero. There is no exception here. In the face of an agentic onslaught, I think you will see that this becomes the case sooner rather than later.
I’m curious, what do you believe the Job-to-be-Done is? To facilitate workshops, or something else (like I do)?
Mike, I found this piece compelling and I agree with the core thesis. The general-purpose infinite canvas is a semantic dead end. A sticky note at a particular set of coordinates means nothing to a machine.
But this problem has already been solved in at least one domain: construction scheduling.
The Last Planner System software category (Touchplan, inTakt, Outbuild, vPlanner) figured this out years ago. They use a constrained canvas. A card is always a task object with defined properties. A line is always a dependency relationship. The ontology is enforced by design.
Result: when the session ends, no human transcription required. The schedule generates automatically because the software knows what everything represents. Some of these tools export directly to Primavera P6 or MS Project with full CPM logic intact.
The difference isn’t canvas vs. database. It’s general-purpose vs. domain-specific.
Miro lets you do anything, so it understands nothing.
Touchplan only lets you do scheduling, so it understands everything you do.
The real question your article raises for me: why hasn’t anyone applied this constrained canvas model to other domains? Strategic planning, product roadmapping, investment decisions. All the things people currently do in Miro with zero semantic output. The template exists. Why isn’t it being replicated?
Hi George, thanks for your comment.. and I typically come up with things that are absolute common sense to me, so I'm not surprised they are being done already. Why isn't it being done in strategic planning? I think it's because the people involved in those functions are not constrained and have focused all of the attention on themselves. It's theater. They'll tell you" they get energy from the process" and my response would be. I don't care if you get energy from it, because you're actually sucking energy out of it. 😄
I believe you’re absolutely right there.
I predict that things are going to change dramatically over the next few years as clever people start developing agentic solutions that drive directly to whatever specific outcome is being sought. I think the agents will be smart enough to adapt to different domains as well
Hey Mike,
I manage the template ecosystem at Mural. I really appreciated hearing your point of view—it was thought-provoking and I found it genuinely interesting. As you might imagine, I disagree with several parts of the argument. A few thoughts after watching your video:
• The goal of a digital whiteboard isn’t efficient data entry. It’s to help people who see the world differently externalize what’s in their heads. Once externalized, those ideas become easier to compare, discuss, and align on. Digital whiteboards are simply the latest extension of a tens-of-thousands-year-old human behavior (e.g., the Caves of Lascaux in France or Altamira in Spain).
• Sticky notes are helpful because they force brevity. Unlike a Google Doc, a sticky note nudges you to distill your thought into a single bullet-sized idea that others can quickly parse. That constraint isn’t skeuomorphic—it’s just a device that helps people communicate more empathetically. (Trello cards work the same way; the “sticky note” metaphor isn’t the important part.)
• Your overall premise seems to be that unstructured whiteboard data is inherently problematic, but I think that reflects a misuse of whiteboards:
a) Your described scenario: Five people generate 100 ideas on a whiteboard, and then all 100 need to be moved into a system of record.
b) How whiteboards are meant to work: A whiteboard helps those five people pull 100 ideas out of their heads, discuss them, narrow them down, and choose the 1–3 that deserve to move forward. Only those selected ideas need to go into a system of record. Most ideas are ephemeral—97 out of 100 shouldn’t live as structured data anywhere. The whiteboard is simply a prop that enables low-pressure collaboration, debate, and decision-making. After years of design work, I’ve come to believe almost every idea belongs in the trash; the job is to find the few that don’t.
• It also seems like you’re viewing digital whiteboards primarily as brainstorming tools. Brainstorming is just one of 100+ activities people do in these spaces. For example, the “jobs to be done” ladder you show at 7:10 works extremely well in a digital whiteboard. (In Mural, there’s a template for this called Whys and Hows, used by thousands of people per month.) My guess is that brainstorming accounts for maybe 10–20% of Mural usage.
• The future you describe at 7:45 makes a ton of sense. Facilitators should only need a few minutes to prep for a session like that—certainly not four hours.
• The “pulling data from systems of record” vision is compelling, but today it requires a huge amount of overhead. Someone has to connect those systems, get approval from a cautious IT team, ensure GDPR compliance, and decide what data should or shouldn’t flow in. It’s still a lot of friction. AI will help lower the burden of installing and maintaining integrations, but for now I’d guess there are tens of thousands of people using sticky notes for every one using a data integration. The scale is currently wildly different.
•Syncing back to systems of record is even harder. In a collaborative tool, it’s difficult to convey clearly what’s being synced and who has control. One person may set up the connection and understand it fully; another might join later and inadvertently change something in Product A, not realizing they’re modifying data in Product B for thousands of users. After 20+ years of designing interfaces, I’d say the level of clarity required here is about as hard as interface design gets. Miro is racing toward this vision, and I think it’s making their product more confusing by the day.
Anyway, thanks again for the thought-provoking video. I enjoyed hearing your perspective, and there’s a lot of wisdom in it—even if I quibble with parts of the argument.
"Sticky notes are helpful because they force brevity" You should really read my article. This is one of several challenges that a first principles analysis would tackle. What is the principle behind brevity facilitating decision-making, and what is specific source. I'll go check it out.
My overall premise is that there is a great amount of waste in the "process" of decision-making. Just like there was a great amount of waste in building rockets. Once the status quo was challenged, we suddenly can get things into space for a fraction of the cost that existed just a few years ago. This is why I recommend reading the entire piece, and maybe some other articles on my blog. I use the exact same FP + JTBD approach to everything.
Ryan, thanks for taking the time. I urge you to read the entire article because the video is a barebones summary.
The point is that the goal ( the outcome) of collaboration is a decision, and workshops are a solution (a means to an end). Facilitators (expensive) are the glue that hold workshops together (and deal with the exact same limitations of a digital canvas).
From an innovation perspective (First Principles + JTBD) this is ripe for disruption in my humble opinion. History has shown us that every industry drives to zero. There is no exception here. In the face of an agentic onslaught, I think you will see that this becomes the case sooner rather than later.
I’m curious, what do you believe the Job-to-be-Done is? To facilitate workshops, or something else (like I do)?