Storyboarding a video ad is straightforward because you control the clock: frame one leads to frame two and the viewer is a passenger. A playable has no clock you control. The user decides when to act, whether to act at all, and how well they do it — so a linear sequence of frames describes one possible playthrough out of many, and usually the most optimistic one.
Storyboard states, not frames
The unit that works is the state: a distinct configuration the ad can be in, plus the events that move it to another state. A minimal playable has four — attract, active, resolved, end card — and most of the value of storyboarding comes from being explicit about the transitions between them rather than from drawing them beautifully.
Written as a list rather than a picture, a simple one looks like: attract loops until first touch; first touch begins active; active resolves on win or on a timeout; resolved plays a beat of feedback; end card appears and persists. That is a buildable specification, and it took one paragraph.
The states people forget to draw
Three of them, consistently. The do-nothing path is first: a meaningful share of impressions produce no interaction at all, and if your storyboard has no state for that, the build will have no behaviour for it — the ad will sit on its attract loop until the container closes and the CTA will never be seen.
The failure path is second. If the player can lose, something has to happen, and "restart" is a decision with consequences — a playable that silently restarts can trap someone in a loop they never escape, which spends your impression on frustration. The third is the interrupted path: the ad becomes non-viewable mid-interaction because the user scrolled or the container backgrounded it. Deciding whether the experience pauses or continues is easy at storyboard stage and awkward afterwards.
The transition worth writing down explicitly
How the resolved state reaches the end card. In an image-editor project the end card is an ordinary scene reached by a Go to Scene action fired from the completion event — not an automatic consequence of the game ending. Storyboards that stop at "player wins" routinely produce builds where nobody ever sees the CTA.
A beat sheet for the first five seconds
Within the active state, the opening still benefits from something closer to traditional beats, because it is the part you do control. Write it in seconds with the interaction available from the first: at 0s the mechanic is on screen and tappable; by 1s a cue points at the first action; by 2–3s the first success has fired and produced visible feedback; by 4–5s the second, harder attempt is underway.
Notice what is not in that list: a logo sting, an intro animation, a tutorial overlay explaining the rules. Each of those spends attention you have not yet earned, and the beat sheet exists mostly to make their absence a decision rather than an oversight.
Two orientations, one board
Draw the landscape arrangement next to the portrait one for at least the active and end-card states. Not because they are different designs — they should be the same design positioned differently — but because doing it at storyboard stage surfaces the elements that have nowhere sensible to go in the other aspect, which is exactly the problem that otherwise appears during review as "it crops through the button".
The practical output is a note per element about what it is anchored to: centre, a safe-area edge, or another element. That is the information a builder needs to make one project serve both, and it is much cheaper to decide with a pen than in a layout tool.
Mark the safe areas
Draw the exclusion zones on the board itself: the notch, the home indicator, and the network's own close button, which typically sits in a top corner and will overlap whatever you put there. Several networks also supply their own close control and ask that the creative not add a second one — worth checking per network, because two close buttons is both a rejection risk and a genuinely confusing experience.
How much fidelity is useful
Very little. Boxes and arrows on paper communicate state machines better than polished mockups, and polish at this stage tends to lock in visual decisions before the interaction has been proven. The failure mode of a beautiful storyboard is that it becomes a specification for pixels rather than for behaviour, and then the build is judged on resemblance instead of on whether the ad works in the hand.
The exception is the end card, which is genuinely a layout problem rather than a behaviour problem, and where a real mockup earns its cost. Everything before it is about timing and response, neither of which a static image can show.
When to stop storyboarding and start building
As soon as the state list is complete and the first five seconds are timed. Interactive creative resists being fully designed on paper — how a mechanic feels at a given difficulty is not knowable until it exists — so the storyboard's job is to eliminate ambiguity about structure, not to pre-empt iteration. With a no-code editor the fastest route to answering "does this feel good" is usually to build the rough version and play it, which is a better use of an afternoon than a fifth revision of the board.
Reviewing a storyboard
The useful review question is not whether everyone likes it but what happens if. Walk the states and ask: what happens if the player does nothing for five seconds; what happens if they fail; what happens if they win in two seconds instead of eight; what happens if the ad becomes non-viewable mid-interaction and then comes back. A board that answers all four is buildable without another meeting.
The second question is where the CTA is in each state. Storyboards routinely draw it only on the end card and leave its status ambiguous everywhere else, which means the builder decides by default — and when the CTA becomes available is one of the highest-leverage choices in the whole creative.
Create playable end cards in minutes—no code required.
Open Playable Ads Maker