See the whole workshop
before anyone enters the room.

Start from the deliverable, not the agenda. Name the thing the session must produce, a decision, a shortlist, a plan, then work backwards: what block produces it, and what does that block need as input, and what block produces that. When the chain reaches the start of the day you have the skeleton. Then design the arc over it: alternate divergence and convergence so ideas get generated and then actually narrowed, put the hardest thinking where energy is highest, and give every block its honest duration including transitions. Finally attach materials to the blocks that use them. On a canvas this whole design is visible at once, which is how you catch the block whose output feeds nothing before the room catches it for you.

An agenda is the visible schedule: what happens at 10:15, who leads it, when lunch is. A workshop design is the machinery underneath: what each block must produce, which later block consumes that output, where the group diverges and where it converges, and how energy is spent across the day. Two workshops can have identical agendas and completely different designs, and only one of them ends with a decision. Docs are good at agendas and bad at designs, because the design is a set of relationships between blocks, and relationships need space to be drawn. That is the argument for planning on a canvas: the second, invisible design becomes as inspectable as the schedule.

It drafts, you design. Describe the goal, the group, and the time available and the AI lays out a credible run of show: blocks on a timeline with purposes, timings, and materials attached. That is a starting skeleton, not a finished session. You then do the actual design work, reordering blocks, tightening timings, wiring deliverables to the blocks that consume them, and the AI helps along the way because it keeps the whole board in context: ask it to fill a gap between two exercises or suggest an opener for a skeptical room. What it deliberately does not do is facilitate. Reading the room and knowing when to abandon the plan stay with you.

You run it on the same board you planned on, and that is the point of the design. The blocks you laid out become the stations of the live session: sticky notes and cards for capture, a frame per breakout group, a marked-out voting area the group drags its strongest ideas into, and a parking lot for tangents. Because the plan and the room are one canvas, nothing gets rebuilt in a second tool the night before, and everything the session produces lands next to the block that produced it, which makes the follow-up write itself. One honest note: Storyflow does not have dedicated voting or timer widgets the way Miro and Mural do; convergence here works spatially, by moving ideas into areas.

It depends on what they need to do. Anyone with a view-only link can open the board in a browser with no account at all, which covers stakeholders, sponsors, and anyone reviewing outcomes after the session. Participants who will actively contribute, adding sticky notes, moving ideas into the voting area, working in a breakout frame, need to be invited to the board as collaborators, and people a paid member invites join free. For a recurring team ritual like a retro that setup happens once. For one-off sessions with external participants, factor a few minutes of onboarding into the arrival block, which is good practice in any tool.

Not during early access. Storyflow is paid-only right now, with plans starting at $9.99 a month, and a Free plan launches before the end of 2026. Anyone a paid member invites to a board joins free today, so your co-facilitator and your participants do not need their own subscriptions. If you need a completely free option this minute, FigJam's free tier is genuinely good: three collaborative boards at no cost, with the voting and timer widgets included, which is enough to run real sessions while you decide whether the planning depth of a paid tool is worth it to you.

Yes, and both benefit from the same treatment. A design sprint is five days of tightly chained deliverables, map to sketch to decision to prototype to test, and laying the whole week out as connected blocks shows you whether each day genuinely feeds the next before you have burned a team's week finding out. A retrospective is the opposite case: a short format you run every sprint, where the win is designing the arc once, from safe check-in through honest divergence to committed actions, then duplicating the board each time and refining it as you learn. Storyflow's template library covers sprint and retro formats, and the AI can lay either out from a one-line description.

Shorter than you think for input, longer than you think for exercises. Talking blocks lose the room after about ten minutes; working blocks almost always need more time than the agenda admits once you count explaining the exercise, the slow start, and the discussion that follows. The practical method is to give every block its honest duration including transitions, add a buffer block you plan to cut, and then look at the whole timeline at once. This is where planning spatially earns its keep: an overstuffed afternoon reads as an overstuffed afternoon when blocks are sized by time on a canvas, whereas a doc happily lets you schedule nine hours of content into six.

Three ways, in increasing formality. Send a view-only link and stakeholders open the live board in a browser with no account, seeing the outcomes in context: the decisions sitting next to the votes and exercises that produced them, which answers the 'why did you choose this' question before it gets asked. Export the board as a high-quality image or PDF when someone needs an artifact for a deck or an email. And for the narrative version, write the debrief in the built-in document editor beside the canvas, with the evidence one glance away instead of in a photo roll. The quiet advantage of running the session on the planning board is that this step needs no reassembly; the outcomes are already where the plan was.