A small brand’s product launch: an AI image use case.

1. Define the assignment and the decision

Imagine an independent brand preparing a cobalt-blue clothbound notebook launch. The founder knows the product and customer; a freelance designer handles the website and social layouts. They need to decide whether a quiet, daylight-led desk setting can support the launch before arranging photography and final design.

The assignment is a concept board with three placements. The founder would approve the mood and check it against the real product. The designer would test composition, crops, and copy space. One person would operate the studio through their own signed-in account and share saved files for review; the current product has no shared project or approval workspace.

• See the studio workflow for small creative teams

2. Fix the visual rules before generating

The team would agree on one visible idea: a calm place to start writing. Keep the rules below in each brief and vary only the frame's job. Consistent instructions provide a review standard; they do not lock the same object across separate generations.

The studio currently takes text prompts and does not accept product-photo uploads. These concepts therefore describe an imagined, unbranded notebook. The founder would keep approved product photographs beside the review board, outside the studio, and reserve accurate product representation for final production.

• Palette: cobalt-blue cloth, warm white surroundings, one graphite accent.
• Light: soft daylight from the left, with gentle shadows.
• Styling: a plain desk and no more than one pencil.
• Copy: quiet space for a headline; add final lettering and logos in the design tool.
• Reject: invented branding, extra accessories, distorted binding, or a different lighting direction.

• Turn a campaign direction into a complete image brief

3. Start with the website hero

Choose Campaign, 3:2 — Landscape, and Medium for the first composition review. The studio's landscape output is 1536 × 1024 pixels. The designer would place a saved candidate into the real page layout and add the draft headline there, so the decision is based on the image's job.

A candidate would pass this stage if the full notebook remains visible in the intended crop, its silhouette reads at the planned display size, and the actual headline sits on a quiet background. An attractive image that leaves no usable copy space would need a revision. The brief below is a starting input, not an example of an observed result.

• Plan a hero image around its headline and crop

4. Explore the material without inventing product facts

Choose Editorial, 1:1 — Square, and Medium for a 1024 × 1024 material study. This frame would help the team discuss light and texture with a photographer. The founder would compare it with a physical sample or approved photograph to identify where the concept differs from the actual cloth and binding.

If a candidate introduces stitching, a ribbon, or an embossed mark the product does not have, exclude that detail from the production brief. A generated surface study cannot establish the notebook's construction. Keep the material reference labeled as a concept, and use an accurate photograph for customer-facing product details.

5. Adapt the idea for the social opener

Choose Campaign, 2:3 — Portrait, and Medium for a 1024 × 1536 concept. Carry forward the palette, light, and single prop, but ask for a taller arrangement. Save the result and test the intended social crop in the design tool; the studio's portrait format is not automatically the final format for every placement.

The designer would compare all three candidates together. A warmer background or harder shadow could make the social image feel unrelated, even if it works alone. Check the proposed crop with the actual copy and keep the notebook away from areas that will be covered or removed.

• Give each image in a social series a clear role

6. Make the next request solve a specific problem

After the first round, the team would review the candidates against the agreed criteria and choose one specific change per next attempt. Save useful versions first. The current studio creates a new image from the revised text; it cannot edit a selected area or guarantee that the rest of a previous image stays unchanged.

The situations below are hypothetical review exercises. If the same critical defect remains after the reserved revision, pause that frame and rewrite the brief or move it into another production workflow. Repeating the same request without a decision would spend the budget without resolving the assignment.

• Write feedback that names the defect and intended change

7. Set the budget and stopping point

For this proposed exercise, set a ceiling of 12 generation attempts and record every attempt, including failures, in the project log. The allocation below is a planning choice, not a prediction of how many images will be accepted. Stop early if the concept board already supports the decision; do not spend the remainder merely because it is available.

The Creator membership is currently advertised at $490 USD per month with up to 120 studio renders. Twelve successful, counted renders would use 10% of that allowance. The membership still costs $490 for the month if this is your only assignment, renews until canceled, and does not include OpenAI API usage. Check current pricing and your remaining allowance before starting.

Budget separately for actual API charges, the designer's time, photography, and final editing. No fixed API bill is assumed here. A fresh High-quality generation is another request and can change the composition; the current studio has no control for upscaling a selected result. Review it again before replacing an accepted concept.

• Check current Render Office membership prices
• Review recurring membership and usage terms

8. Hand off the decision with the files

Save each useful image with the browser's image-save option before another successful generation replaces it or the page is reloaded. The current studio displays one result and has no saved-project history. Keep the shared review board and production notes in the team's existing tools.

The intended handoff is a direction the founder has approved for production, with unresolved details stated clearly. It is not a claim that three sale-ready product photographs have been created. The photographer and designer would use the selected mood, framing, and lighting while making final assets from the real product.

• Three saved concept files, named by placement and version, such as notebook-hero-concept-v02.png.
• The exact prompt, mode, format, and quality beside each file; keep API keys out of shared notes.
• A draft layout showing the intended crop, final-copy position, and any placement constraints.
• A production note separating approved art direction from product details that need accurate photography.
• The founder's decision, the designer's open issues, and the next owner for each deliverable.

• Prepare a campaign handoff with assets and approval notes

9. Measure whether the exercise earned its place

The proposed success condition is three usable concept directions with placement checks and production notes, approved within the agreed attempt ceiling. That target has not been tested here. A partial set or a decision that the workflow does not fit is a result to record honestly, rather than turning it into a success claim.

Before starting a real assignment, record the hours and full cost of a comparable previous concept-development job, if one exists. Then record generation attempts, review and editing hours, actual spend, accepted concepts, and unresolved issues for this exercise. If there is no comparable baseline, leave it unknown. Generation time alone would not establish a saving across the complete workflow.

For this fictional team, the studio is worth evaluating when the open decision concerns a campaign's mood, lighting, and composition. If the immediate deliverable requires exact product reproduction, persistent image editing, or a shared approval system, those requirements call for another workflow. A useful concept board earns its place by helping the team make a production decision; sales performance would need separate evidence after a real launch.

• Check setup requirements before creating your first image