⚡ Promptolis Original · Creative & Arts
🧪 Midjourney Ideation-to-Production Pipeline
Rapid ideation with --chaos + --weird, then lock aesthetic with --sref for production.
Why this is epic
Most Midjourney users either (a) iterate randomly hoping to stumble onto something good, or (b) produce one-shot images that work individually but don't compose into cohesive projects. Professional workflow: deliberate ideation (--chaos + --weird for surprise) → aesthetic lock (--sref anchor) → production (consistent follow-ups). This prompt walks you through the full pipeline.
Takes your project brief + timeline + end-deliverable → outputs a 3-stage plan: Ideation (8-15 images, high variation), Aesthetic Selection (how to pick the anchor from ideation), Production (--sref-driven image set). Also covers when to REJECT an ideation direction vs. refine it.
Handles the common mid-project problem: aesthetic drift across 10+ images. The pipeline's job is preventing drift through early-locked anchor, not fixing drift after 20 images are already generated.
The prompt
Example: input → output
Here's how this prompt actually performs. Real input below, real output from Claude Opus 4.
<project-brief>Creating 12-image series for a personal finance newsletter's year-in-review issue. Each image illustrates a specific financial concept (saving vs investing, emergency funds, compound interest, lifestyle inflation, etc.). Need to feel editorial + slightly approachable + not-intimidating. Delivery Friday (4 days).</project-brief> <aesthetic-direction>Editorial illustration, muted palette, conceptual not literal, feels like New Yorker-adjacent but slightly warmer/more approachable</aesthetic-direction> <project-scale>Medium 12 images</project-scale> <timeline>4 days</timeline> <midjourney-version>v7</midjourney-version>
Project Pipeline Plan
4 days is tight but workable for 12 editorial images. Proposed split: Day 1 = ideation (3 directions, 4 images each). Day 2 morning = anchor selection + 3 production test images. Day 2 afternoon through Day 3 = production of 10 remaining. Day 4 = revisions + buffer. Tight but achievable with disciplined pipeline.
For newsletter year-in-review: conceptual > literal is right (editorial register). 'Saving vs investing' as concept needs conceptual illustration (scales, growing seed, bridge), not literal (piggy bank, stock ticker). We'll lock this register in ideation.
Stage 1: Ideation (Day 1, 3-4 hours)
Generate 3 direction explorations, 4 images each = 12 ideation images. Each direction explores a different aesthetic angle within your brief.
Ideation Prompts
Direction A — Warm Editorial (primary candidate)
Editorial illustration showing conceptual personal finance concept, limited palette of muted ochre cream and slate blue, flat bold simplified shapes, mid-century editorial register, slightly warmer than New Yorker, off-center composition with negative space, serious but approachable tone --ar 4:3 --stylize 600 --chaos 50 --v 7Direction B — Slightly Playful
Playful editorial illustration for personal finance concepts, limited warm palette of terracotta cream sage and dusty blue, flat geometric shapes, slightly whimsical conceptual framing, modern editorial register like Guardian or Economist, approachable composition --ar 4:3 --stylize 700 --chaos 50 --weird 250 --v 7Direction C — Minimal Nordic
Minimalist Scandinavian editorial illustration for financial concepts, muted palette of soft beige sage and slate, flat bold shapes with heavy negative space, contemplative mood, simplified conceptual metaphors, limited detail --ar 4:3 --stylize 500 --chaos 50 --v 7Generate each prompt 4 times (via Midjourney's 4-grid system = 16 images per prompt). Total: 48 images across 3 directions.
What to Look For
Selection criteria by end of Day 1:
1. Best-fit to brief: Which direction feels most 'editorial + approachable + not intimidating'?
2. Productionability test: Pick one direction. Imagine 12 different finance concepts illustrated in that direction. Does the aesthetic survive across concepts, or does it break down for specific subjects?
3. Distinctiveness: Which direction is different enough from generic 'editorial illustration' to feel like YOUR newsletter's voice, not generic?
Red flags in ideation:
- Direction where all 4 generations look nearly identical (too narrow for 12-image series)
- Direction where outputs feel generic/stock-illustration (not distinctive enough for branded use)
- Direction where register feels too playful OR too serious (contradicts 'approachable + editorial' brief)
My guess: Direction A (Warm Editorial) likely wins for this brief. But generate all 3 before deciding.
Stage 2: Aesthetic Lock (Day 2 morning, 2-3 hours)
Anchor Selection
From your winning direction, pick the single strongest image as anchor. Criteria:
1. Most representative of the direction (not the outlier)
2. Composition has room to vary (if it's too specifically composed, follow-ups will feel forced into same composition)
3. Palette is clean + consistent (not an accidental muddy render)
4. Subject is abstract enough (avoid anchoring on image that's TOO specifically about one finance concept — you need flexibility for 12 different concepts)
Often your 'favorite' image isn't the best anchor. Your favorite might be one-off magic; the best anchor is the most reliably-reproducible representative of the aesthetic.
--sref Workflow Setup
Once anchor is picked:
1. Copy anchor URL (save in Notion or project doc)
2. Generate 3 test production images using --sref to validate the anchor holds across different subjects:
[ANCHOR_URL] Conceptual illustration of compound interest growing over time, editorial illustration style, limited palette ochre cream slate, flat bold shapes, off-center composition --ar 4:3 --style raw --sref [ANCHOR_URL] --sw 100 --stylize 600 --v 7[ANCHOR_URL] Conceptual illustration of lifestyle inflation trap, editorial illustration style, limited palette ochre cream slate, flat bold shapes, off-center composition --ar 4:3 --style raw --sref [ANCHOR_URL] --sw 100 --stylize 600 --v 7[ANCHOR_URL] Conceptual illustration of emergency fund concept, editorial illustration style, limited palette ochre cream slate, flat bold shapes, off-center composition --ar 4:3 --style raw --sref [ANCHOR_URL] --sw 100 --stylize 600 --v 7If all 3 test images feel like siblings + match anchor aesthetic: proceed to production.
If 1-2 test images drift: adjust --sw (try --sw 150 for tighter consistency, --sw 75 for more flexibility). Test 3 more.
If all 3 test images feel off or vary wildly: anchor is wrong. Return to ideation images + pick different anchor. Don't force.
Stage 3: Production (Day 2 afternoon - Day 3, per-image ~10-15 min)
Production Prompt Template
[ANCHOR_URL] Conceptual illustration of [SPECIFIC_FINANCE_CONCEPT], editorial illustration style, limited palette ochre cream slate, flat bold shapes, [COMPOSITION_VARIATION for this concept], approachable serious tone --ar 4:3 --style raw --sref [ANCHOR_URL] --sw 100 --stylize 600 --v 7For each of the 12 concepts, fill in:
- [SPECIFIC_FINANCE_CONCEPT]: the concept being illustrated
- [COMPOSITION_VARIATION]: 'off-center left' / 'diagonal asymmetric' / 'triangular arrangement' / etc. — varies composition subtly for visual variety within aesthetic
12 concepts to illustrate:
1. Compound interest (growing element over time)
2. Emergency fund (protective/buffer metaphor)
3. Saving vs investing (two paths / two tools)
4. Lifestyle inflation (expanding outward)
5. Debt avalanche vs snowball (melting / rolling)
6. Retirement timeline (distant horizon)
7. Tax efficiency (sieve / filter)
8. Diversification (scattered vs. concentrated)
9. Risk tolerance (climbing / descending)
10. Behavioral finance (split internal figure)
11. Opportunity cost (fork in path / abandoned alternative)
12. Year-in-review reflection (retrospective pose)
When to Return to Ideation
If by image 5-6 of production you're consistently getting outputs that feel off:
- Maybe the anchor was wrong (return to ideation, pick different anchor)
- Maybe --sw needs adjustment (try different strength)
- Maybe the 12 concept choices don't all fit this aesthetic (some concepts may need hybrid real-photo + illustration approach)
Don't force 12 when 9 are working. Better to deliver 9 strong + 3 photographs or typography covers than 12 mediocre illustrations.
Project Timeline Breakdown
- Day 1 (Monday): Ideation — 48 images across 3 directions (3-4 hours)
- Day 2 morning (Tuesday AM): Anchor selection + 3 test production images (2-3 hours)
- Day 2 afternoon + Day 3 (Tuesday PM through Wednesday): Production of 10 remaining images (~1 hour per image, 10-12 hours total)
- Day 4 (Thursday): Revisions + buffer (3-4 hours)
Budget some Day 4 buffer for 2-3 images needing full re-do (common in medium-series work). If no re-do needed, use buffer time for background-removal or post-processing.
Caution: This timeline assumes dedicated focus time. With interruptions, double estimates.
Common use cases
- Art directors planning cohesive image sets for campaigns / publications / books
- Content creators scaling visual output while maintaining brand consistency
- Brand designers establishing new aesthetic systems for rebrands
- Photographers exploring AI-augmented workflow for client projects
- Independent publishers producing cover art series
- Freelance illustrators working efficiently on client deliverables
Best AI model for this
Claude Opus 4 for project-planning; Sonnet 4.5 for tactical execution.
Pro tips
- Ideation time = 20-30% of project time. Don't skimp. Your locked aesthetic comes FROM ideation; shortcutting ideation = shallow aesthetic decisions.
- --chaos 50-75 for initial ideation. --weird 500-1000 for creative surprises. Neither for production; production uses --chaos 0 for consistency.
- Selection criteria for anchor: (a) best-fit to brief, (b) productionable (can you generate 20+ variants without running into repetition), (c) distinctive enough to be brand-identifiable.
- Don't anchor on your FAVORITE ideation image. Anchor on the image that represents the most productionable DIRECTION. Sometimes favorite is a one-off magic.
- After anchor lock, if 3+ production images still feel off — go back to ideation. Don't force production from wrong anchor.
- Keep an 'ideation library' of rejected-but-interesting images. Future projects may use these as references for different contexts.
Customization tips
- For ongoing newsletter work: after year-in-review, keep the anchor URL for future newsletter editorial needs. Aesthetic consistency across issues = brand building. Same anchor across 3-6 months of editorials.
- For rapid-deadline work (24-48 hours): compressed pipeline. 1 hour ideation, 30 min anchor lock, rest is production. Skip Direction exploration; pick one direction based on quick judgment.
- For stakeholder review processes (agency, brand work): surface 3-5 anchor candidates (not the full 48 ideation images) + 3 test production images per candidate. Stakeholders approve the anchor; production proceeds from locked anchor.
- For ongoing client work (repeat illustrator contracts): build anchor library for different client aesthetics. Each client gets 2-3 anchor URLs that represent their brand. Work is faster over time as anchors are reused.
- For experimental/art projects where aesthetic-discovery IS the project: skip Stage 2 aesthetic lock. Stay in ideation mode longer. Different workflow; different output (less commercial, more personal).
- For team projects: share anchor URL + --sw + template prompt as team documentation. Everyone generating production images works from same anchor, same template. Prevents drift across team members.
- For learning the pipeline as beginner: start with small projects (3-5 images). Master the pipeline on small scope before attempting book-length work. Pipeline habits transfer.
Variants
Default Pipeline
Standard 3-stage workflow for 5-20 image projects
Single-Image Project
Compressed pipeline — rapid ideation (3-5 images), direct production
Book-Length (30+ images)
Extended ideation + aesthetic lock + production phase management
Brand Rebrand
Aesthetic exploration before commitment, stakeholder review checkpoints
Rapid Deadline
Compressed pipeline for tight-timeline work — ideation in 30 min, anchor same day, production next day
Frequently asked questions
How do I use the Midjourney Ideation-to-Production Pipeline prompt?
Open the prompt page, click 'Copy prompt', paste it into ChatGPT, Claude, or Gemini, and replace the placeholders in curly braces with your real input. The prompt is also launchable directly in each model with one click.
Which AI model works best with Midjourney Ideation-to-Production Pipeline?
Claude Opus 4 for project-planning; Sonnet 4.5 for tactical execution.
Can I customize the Midjourney Ideation-to-Production Pipeline prompt for my use case?
Yes — every Promptolis Original is designed to be customized. Key levers: Ideation time = 20-30% of project time. Don't skimp. Your locked aesthetic comes FROM ideation; shortcutting ideation = shallow aesthetic decisions.; --chaos 50-75 for initial ideation. --weird 500-1000 for creative surprises. Neither for production; production uses --chaos 0 for consistency.
Explore more Originals
Hand-crafted 2026-grade prompts that actually change how you work.
← All Promptolis Originals