Step 01
Build a Mood Board
Collect images in Pinterest or Cosmos that express the color, lighting, texture, composition, and attitude you want. Keep the first board coherent enough to test, but do not treat it as final.
IntermediateDesign
How to Create a Consistent Brand Aesthetic in Midjourney
Build a repeatable Midjourney brand system by testing a visual mood board, isolating the style references that actually work, layering in a personalization code, and documenting the final recipe for reuse.
Try this with your AI assistant
Download the skill for this workflow. Version 1.0.0.
Source episode · 03:00How I AI: Jamey Gannon's Workflow for Consistent Brand Imagery in Midjourney
with Claire Vo
Jamey moves from a Pinterest mood board to individual style references, removes references that distort the palette, adds a personalization code, and packages the final recipe in Figma.
Before you start
What you need
- A focused mood board
- A Midjourney account
- Three to five representative test subjects
- A Figma file for the final system
What you’ll make
A documented visual recipe that produces a recognizable family of brand images across different subjects.
Tools used
- MidjourneyVisitMidjourney
AI image generation
- FigmaVisitFigma
Collaborative design tool
The workflow
Follow the sequence once, then adapt the prompts, checks, and handoffs to your own setup.
5 steps
Step 02
Test with Individual Style References (SREFs)
Run the same small set of test subjects with individual mood board images as style references. Compare the outputs, identify which reference drives an unwanted color or treatment, remove it, and rerun.
Example prompt
Run these subjects with the attached style references: ethereal female model, astronaut, cat, and runner. Keep the subject prompts unchanged so I can compare what each reference contributes to color, lighting, texture, and composition.
Step 03
Layer in a Personalization Code
Add a Midjourney personalization code that reflects the taste you want to layer onto the reference set. Compare results with and without the code, and keep it only when it improves consistency across the test subjects.
Example prompt
ethereal female model --sref [reference URLs] --p [personalization code]
Step 04
Use Simple 'Cheat Code' Prompts and Smart References
Use short subject prompts plus compact visual signals such as a publication, camera, material, or composition reference. Crop distracting objects out of image references when Midjourney keeps copying them.
Example prompt
Dazed editorial photo shoot, model on a matte black leather couch, Sony RX100 --sref [reference URLs] --p [personalization code]
Step 05
Package the Aesthetic for Clients
Package the working recipe in Figma with the exact style references, personalization code, prompt pattern, settings, example subjects, and representative outputs. Note where a separate subject-specific reference set is needed.
What good looks like
- The same reference set works across several test subjects.
- Removing one reference has an explainable effect on color or composition.
- The final images share a visual through line without becoming identical.
- Another person can reproduce the look from the packaged references and settings.
Build your next product with ChatPRD
Turn an idea into a PRD, user stories, and a plan.
Runbook notes
How to recover when the loop fails and where human judgment helps.
If it goes sideways
- The mood board averages into a generic look
- Use individual style references, remove the image pulling the output off course, and rerun the same test prompts.
- A personalization profile adds an unwanted medium or motif
- Create a narrower profile, skip ambiguous choices, and compare it against the same reference set without personalization.
- The style works for people but not products or animals
- Add references that pair the desired aesthetic with the missing subject class instead of forcing one reference set to cover everything.
- A client cannot reproduce the results
- Record the exact references, profile code, prompt pattern, stylization settings, and example outputs in one Figma handoff.
Related workflows
intermediateEngineeringMeta MuseClaude Code
How to Set Up Meta's Muse Wonder for Headless AI Testing
A step-by-step guide to signing up for Meta's Muse API, installing the Muse Code command-line tool, and running it headlessly with an external agent like Claude Code for easier testing and benchmarking.
Open
advancedOperationsMeta Muse
How to Test an AI Agent for Complex Customer Support Ticket Resolution
Benchmark an agentic AI model's ability to resolve complex customer billing issues. This workflow tests the AI's ability to investigate tickets, handle complex logic, execute changes in an admin app, and communicate with the customer.
Open
intermediateProductClaudeGoogle Drive+1
Generate New Video Ideas by Analyzing Your TikTok Comments with Claude
Use your existing TikTok video library and their associated comment threads to generate highly relevant new content ideas and scripts. This workflow pairs audience feedback directly with the specific video performance that prompted it, leading to targeted and engaging new topics.
Open
Start shipping better products.
Join 250,000+ product managers who use ChatPRD to write better docs, align teams faster, and build products users love.
Free to start · 14-day Pro trial · SOC 2 Type II
