Storyboard a Short Product Ad Before Generating Video

Start with a message the product facts can support

Imagine a rigid, sage-green desk tray with three fixed compartments. Its intended audience wants a defined place for small everyday desk items. The ad's message is “Give small things a place.” It should show the product early, explain its layout, reveal a useful detail and end with a simple invitation to explore it.

These facts are assumptions inside this fictional brief, not specifications for a real brand or SKU. A real project would need approved product photography and a checked fact sheet before those details could be accepted.

The dimensions are this example's delivery choice, not a universal advertising-platform specification. The props are illustrative and are not included with the tray. Keep that distinction visible in the brief and any required qualification.

If you cannot yet decide the single message or how the picture should express it, use the marketing-brief concept guide first. Here, the next job is to turn a chosen product message into an inspectable sequence with production instructions.

Give each shot one information job

The four beats are context, arrangement, detail and choice. They are deliberately different. A second attractive hero view would add little if it did not explain something the first view missed.

The total is 3 + 4 + 4 + 4 = 15 seconds. No extra title or transition time has been added. If an end card is inserted later, something else must shorten or the target must change. A three-second editorial selection is not necessarily a three-second generation option; record those as separate quantities.

Original planning schematic, not product photography, generated footage or a tested advertisement. Each panel needs its own approved asset before production.

The arranged-items shot is a cut to a different state, not a demonstration of the tray sorting objects automatically. No hand needs to place several items while the camera circles. A simple board makes the product explanation easier to review and leaves fewer physical events for a future generation to invent.

Card 1: Establish the product and situation

Purpose and framing: A three-quarter desk view shows the whole tray from the start. A few small items sit nearby. The product should not be hidden until the last instant of the opening; this brief wants immediate recognition.

Required asset: An approved three-quarter photograph showing the correct outer shape, sage-green color and visible compartment layout. The source must have enough detail for the planned crop. A mood image can guide the desk setting but cannot define the tray's geometry.

Start and end: Begin with the tray clearly visible and the example items outside it. End with the same arrangement, allowing enough time to read the opening line. There is no packing action or before-and-after claim in this shot.

Camera and motion intent: Use a locked composition or a very small proposed push toward the product. Preserve room for the headline and keep the full required product inside the frame. A future generation must be inspected for unwanted shape changes; the camera instruction is not a guarantee.

Copy and sound: Add “Give small things a place” as editable text in the finishing layout. The rough board can be silent. If room tone or music is considered later, it should support the sequence rather than supply a missing product explanation.

Proposed starting-image instruction:

Use the approved three-quarter photograph of the sage-green three-compartment tray as the product authority. Plan a clean portrait desk composition with the whole tray visible and a few approved example items beside it. Preserve the tray's actual outer shape, color and divider layout. Leave a clear headline area above. No lettering, added compartments, extra tray or implied automatic sorting.

Proposed future motion instruction:

One continuous desk shot with a restrained camera push toward the supplied tray composition. Keep the tray and nearby props in place and recognizable. Keep the whole required product within the frame and the headline area clear. No hand interaction, object transfer, new product feature or cut.

These are unexecuted teaching instructions. For a real run, choose an available generation duration that can supply the needed usable interval; the board retains three seconds, rather than assuming the interface accepts a three-second request.

Reject or revise when: The product is obscured, a divider changes, the push crops the edge, or the opening requires more reading time than allocated. Fix the asset or board before treating a later attractive clip as a solution.

Card 2: Explain the compartment layout

Purpose and framing: Cut to a top-down view that makes the three compartments easy to count. Place the approved example items in distinct spaces. This shot explains structure, rather than repeating the opening atmosphere.

Required asset: A real approved top-down product view, plus a checked arrangement of props. If the source shows only the outside, obtain the needed view. Generating an attractive unseen interior does not establish the actual product layout or fit.

Start and end: Begin with the items already arranged. Hold that state long enough for the three-part organization to be understood. The cut from the first shot indicates a new presentation view; it does not depict the missing placement action.

Camera and motion intent: A stationary top-down composition is sufficient. A subtle emphasis may be proposed later, but there is no need for rotating objects or a moving hand. The product and prop count stay fixed.

Copy and sound: Use the approved label “Three compartments” as native type. Avoid “fits everything” or a capacity claim. Do not add a drop or click sound that suggests an action the picture never shows.

Proposed starting-image instruction:

Use the approved top-down product photograph and approved prop arrangement. Plan one portrait composition showing the same sage-green tray and its three fixed compartments clearly. Keep the actual dividers and proportions. Show only the verified example items in their planned spaces. Leave a clear area for a short feature label. No text, extra compartment, duplicate item or invented interior.

Proposed future motion instruction:

One stationary top-down shot of the supplied arranged tray. Keep the product, three compartments and example items in the same positions. Allow only restrained ambient visual movement if appropriate. No packing action, camera orbit, extra object or cut.

Reject or revise when: A compartment becomes hidden, props no longer fit, the product appears to have more than three spaces, or the arrangement implies contents that are actually sold separately. If a fit is unverified, remove that prop or obtain supporting photography.

Card 3: Show a detail that helps identify the product

Purpose and framing: A closer view shows the divider junction and enough outer structure to locate it within the tray. It gives the viewer a clearer look at the actual design without turning texture into an unsupported durability claim.

Required asset: A supplied detail photograph matching the previous views. Do not use a generated seam, coating or fastening as proof that the real tray has it. Choose a detail the source actually reveals.

Start and end: Begin on the divider junction and keep that feature readable through the selected interval. End with the same recognizable structure. There is no requirement for a new side or an unseen underside to appear.

Camera and motion intent: Use a static detail or a small proposed push. Avoid an orbit around hidden geometry. Keep the feature within the crop and the same light direction as the preceding shots where the sources allow it.

Copy and sound: This shot can carry no additional words. “Three compartments” has already communicated the factual feature. Adding “premium,” “indestructible” or a strength demonstration would require a different factual brief and evidence.

Proposed starting-image instruction:

Use the approved divider-detail photograph of the same sage-green tray. Plan a portrait close view that keeps the divider junction and part of the outer rim legible. Preserve actual geometry, surface and color. Match the restrained desk treatment. No new texture claim, invented fastening, writing or additional product.

Proposed future motion instruction:

One restrained detail shot with a slight push toward the visible divider junction. Keep its geometry, rim and surface recognizable. Do not reveal an unsupplied underside, change the material, bend the divider or introduce a cut.

Reject or revise when: The detail could belong to a different product, the macro crop loses its relationship to the tray, or the camera reveals invented construction. A static approved photo remains an option if motion adds risk without useful information.

Card 4: Make the final choice clear

Purpose and framing: Return to a clean hero view with space for the product name and call to action. Keep the product proportional and make the copy easy to read in the selected portrait format.

Required asset: The accepted hero photograph or a checked composition derived from it, plus approved product naming and the real destination supplied by the project. Do not invent a live URL for the teaching example.

Start and end: Begin on the final layout and hold it. There is no new action competing with the call to action. The ending must remain visible through the full intended four-second selection.

Camera and motion intent: A still hold is acceptable. If the project chooses motion, keep it small enough that the copy area and important product features remain useful. The final asset need not be generated video simply because earlier shots are moving.

Copy and sound: Use “Explore the tray.” Add the approved product name and, if required by the real brief, “Tray only; props not included.” Keep those as separately revisable text. Plan a deliberate audio ending; do not assume a generated soundtrack will stop neatly at the final frame.

Proposed starting-image instruction:

Use the approved hero product image to plan a clean portrait closing composition. Keep the same rigid sage-green three-compartment tray clearly visible. Leave room above for the supplied product name and below for the call to action and any required qualification. No generated text, extra accessories, price badge or unapproved claim.

Proposed future motion instruction, if motion is needed:

Hold the supplied closing composition with the tray stationary and fully visible. Keep the designated text areas clear. No new prop, camera orbit, product transformation or cut. The final copy will be added separately in the editor.

Reject or revise when: The text needs more time, a destination is missing, the product changes, or a moving foreground competes with the action line. A visually quiet ending can be the correct choice.

Build and time the board before generating motion

Draw each panel on a separate page at the intended shape. Rectangles, arrows and labeled copy boxes are enough for this stage. Record the asset filename, shot purpose and timing under each page. The combined teaching diagram above explains the plan; it is not one scene's first-frame input.

First, review the four pages in order without sound. Show them for three, four, four and four seconds using a timer or a simple timed presentation. Read the copy at a realistic viewing size. Ask whether the product, its structure and the final action are understandable without someone explaining the board aloud.

For a digital animatic in Premiere, first set Still Image Default Duration to four seconds in Timeline preferences, before importing the four individual panel images. Adobe documents this still-image duration setting. Import the panels into a project, create a sequence and place them consecutively on its timeline, using the documented sequence and media route. Trim the first panel to three seconds and close the resulting gap. Adobe documents edge trimming and gap removal.

Play the resulting fifteen-second rough board and inspect the actual boundaries. This is a documentation-supported proposal, not a completed software test. It sequences still images; no AI-generated motion is needed to judge the information order and reading time.

If the feature label is unreadable, shorten or reflow the wording, enlarge it, or rebalance the durations. For example, allowing one extra second for the closing qualification means removing one second elsewhere or changing the total target. Do not silently add time after approving a fifteen-second brief.

If voiceover is proposed, record its intended words separately and read them against the board. A line describing rapid packing would add a claim the example does not support. Music can suggest pace, but it cannot prove product fit or make missing information understandable.

Prepare optional style frames without confusing their status

Once the rough sequence works, you may want a more specific visual treatment. Dreamina's current Seedream 5.0 Pro guide describes AI Image, text or image upload, Image 5.0 Pro, available aspect and resolution settings, generation, review and download. Use a card's image instruction to create a candidate style frame if the project needs one.

Keep its status clear. A style frame can help approve composition, light and atmosphere. It does not become an authoritative product photograph merely because it looks realistic. Compare the product with the supplied view, check the compartment count and reject invented geometry before it enters a motion request.

URL-based ad tools also need product information. Creatify's URL-To-Video documentation describes using a product URL or manually uploaded assets and information. However those inputs arrive, verify them against the brief. Automatic ingestion does not settle which claim, view or sequence is appropriate.

When a generation route supports references, assign their jobs explicitly: actual product appearance, scene composition or style. A labeled four-panel board is not interchangeable with an individual starting image. For later motion, the current Seedance 2.5 how-to documents the Omni reference route and asset references; a dedicated first-frame control has a different role. Choose the available mode after the board is accepted.

Make revisions propagate to the right shots

A board should tell you what a change affects. If the sage-green product changes to a different approved color variant, recheck every product image, detail, style frame and copy that names it. A single updated hero image cannot repair the old macro view automatically.

If only the call to action changes, revise the closing copy and destination, then test the new reading time. The product shots may remain useful. If the compartment count changes, the feature label, top-down view, detail view and underlying brief all need review. That is a product-fact change, not a cosmetic edit.

Keep the old version identifiable and name the new one. Mark which assets are accepted, which are candidates and which are missing. When a later operator receives the board, they should be able to select the exact input for each shot without guessing.

Hand over a shot-ready plan and stop at the approval decision

The complete package contains the fact sheet, four timed cards, separate panel images, actual approved product sources, copy sheet, sound intentions and unresolved questions. Record whether the sequence is accepted for motion production or still held for an asset or message decision.

For this example, the plan is usable as a teaching board, while actual product photographs and a real destination remain required for production. No generated shot or finished-ad approval is claimed. A real project should close those dependencies before submitting product imagery as factual evidence.

If the same message becomes a swipeable format, rethink the information per card rather than exporting video frames and calling them a carousel. The product-launch carousel guide develops that separate format. For the next stage of motion generation, editing and revisable finishing, use the complete video workflow guide.

Start with rough panels in Dreamina or a simple drawing tool. Approve what the viewer learns, which actual assets support it and how the fifteen seconds are allocated. The resulting storyboard gives generation a specific task—and gives review a clear reason to accept or reject each shot.