5 steps to generate consistent brand images with Midjourney

2026-03-09 · Source: How I AI · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, AI Creative Production · Depth: Intermediate, extended

Summary

Jamie Ganon, an AI creative director, outlines a detailed workflow for generating consistent, high-quality brand assets using Midjourney, Nanobanana, and other tools. The process begins with creating mood boards on platforms like Pinterest or Cosmos, which are then used as Style References (SRFs) in Midjourney to establish a visual aesthetic. Ganon emphasizes iterating with personalization codes and advanced prompting techniques, such as referencing specific magazines (e.g., "Days editorial photo shoot") or camera models (e.g., "Sony RX100"), to refine outputs. The workflow also includes using image references for composition and leveraging tools like Nanobanana for upscaling and precise editing, such as replacing computer screens or fixing hands. The goal is to create a repeatable system that allows clients to generate their own brand-consistent imagery, shifting the creative director's role to upfront definition and code provision.

Key takeaway

For product designers or creative technologists aiming to establish a consistent brand aesthetic with AI, focus on defining a robust visual language upfront. Utilize mood boards and Midjourney's Style References (SRFs) to guide initial generations, then refine with personalization codes and specific stylistic prompts like camera types or editorial styles. This approach minimizes extensive text prompting and empowers clients with a repeatable system for on-brand asset creation, fostering a more collaborative service model.

Key insights

Consistent AI image generation relies on structured workflows, visual language, and iterative refinement.

Principles

• Visual references are powerful prompts.
• Iterate quickly to gather information.
• Combine styles for unique aesthetics.

Method

Start with mood boards for SRFs, iterate with personalization codes and advanced prompts (e.g., magazine styles, camera types), then use tools like Nanobanana for upscaling and precise edits to achieve consistent brand assets.

In practice

• Use Pinterest/Cosmos for mood boards.
• Experiment with Midjourney's SRFs.
• Crop images to remove unwanted elements.

Topics

• AI Creative Direction
• Midjourney Workflow
• Generative AI Brand Assets
• AI Image Editing
• Prompt Engineering Techniques

Best for: Prompt Engineer, Product Designer, Creative Technologist

Related on AIssential

• ChatGPT Images 2.0 First Impressions — AI tools can generate and iteratively refine brand kits based on textual and visual inputs.
• A picture is worth a thousand words for an LLM — Mood boards and style references are crucial for guiding AI image generation towards a specific aesthetic.
• Remaking Old Computer Graphics With AI Image Generation — AI image generation can remake old graphics, but requires iterative prompting and tool-specific techniques.
• The Design Mode for Claude Code... — Achieve pixel-perfect AI-generated UI designs by providing comprehensive context beyond screenshots.
• Accelerating your marketing ideation with generative AI – Part 2: Generate custom marketing images from historical references — Integrating historical campaign data with generative AI enhances marketing content creation and ensures brand consistency.
• Midjourney, known for AI image generation, unveils a full-body ultrasound scanner and its own spa — Midjourney, an AI image firm, is surprisingly diversifying into medical hardware and wellness spas.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by How I AI.