Design
Product
Ecommerce
Merchandising
January 20, 2026
Mandisa Foster

Stop Doing This: The 7 Most Common Prompting Mistakes in Raspberry AI

Why good prompts are essential for great outputs.

Prompting is part art, part science - and like any craft, a few small missteps can throw off your whole result. Maybe you’ve asked for “a trendy dress” and gotten something totally different from what you pictured. Or you’ve packed so many ideas into a single prompt that the AI doesn’t know which one to prioritize. In this article, we’ll break down the most common prompting mistakes we see from designers, why they lead to off-target results, and how to fix them so you get outputs you’re excited to use.

Mistake #1: Overstuffed Prompts

  • Cramming too many conflicting ideas into one.
  • Your prompts can be short and sweet, or long and detailed. Be sure to check that there’s no conflicting information in your prompt.

Bad prompt

Oversized fitted cropped long jacket with minimal detailing but bold graphics, sporty yet luxury, technical but vintage-inspired, neutral colors but high contrast.

Why it fails

The prompt contains conflicting instructions (oversized and fitted, cropped and long, minimal and bold), forcing the model to guess.

Step 1 Sketch to Render prompt

Oversized cropped utility jacket with a relaxed silhouette in black nylon.

Step 2 Edit Module prompt

Initial Prompt

Add bold graphic paneling in black, white and teal with high-contrast detailing. Prompt Assist prompt: Adorn this black utility jacket with bold, high-contrast geometric graphic paneling in black, white, and teal, precisely integrated across its main body, sleeves, and collar. The paneling should feature sharp lines and shapes, creating a modern, dynamic visual while meticulously preserving the jacket's original relaxed silhouette, four flap patch pockets, structured collar, and all brass button details exactly as depicted.

Mistake #2: Vague Language

  • Using terms like “cool” or “pretty” without detail.

Bad prompt

A cool, modern utility jacket.

Better prompt

Modern utility jacket with a clean silhouette and minimal detailing.

Great prompt

Oversized utility jacket with a boxy silhouette, matte olive nylon fabric, clean seam lines, and minimal brushed gold hardware.

Why this prompt is great

The prompt describes what makes the style modern: the clean seam lines and minimal hardware, without using vague terminology.

Mistake #3: Missing Key Visual Details

  • Forgetting to include garment type, style, color, fabrication.

Bad prompt

Utility jacket.

Why it fails

The AI is missing critical information.

Better prompt

Oversized utility jacket in olive green nylon with large front patch pockets and a structured collar.

This works because it clearly defines the product, color, material, and style.

Mistake #4: Underuse of Reference Images

  • Reference Images are a powerful tool to guide your outputs. Don’t rely on text only when you want a specific vibe.
  • When using reference images in tools such as Lifestyle Photography and Product Photography, play around with lower Similarity Slider percentages.

Bad prompt (text only)

Oversized utility jacket with a technical, modern feel.

Better approach

  • Upload a reference image of a utility jacket with the desired silhouette and pocket structure
    Prompt: Oversized utility jacket with a technical, modern feel, structured pockets, and clean seam lines.
  • Similarity slider: 30–45%.

Mistake #5: Not Matching Terms to the Tool

  • Writing On-Body prompts like Lifestyle prompts.

Bad (On-Body)

Model walking through an industrial city at sunset, cinematic lighting, editorial mood.

Why it fails

On-Body prompts should describe how the garment is worn, not the scene.

Better (On-Body)

Model wearing an oversized utility jacket unbuttoned over shirt.

  • Writing Sketch to Render prompts like Edit Module prompts.

Bad (Sketch to Render)

Make this jacket oversized and boxy.

Why it fails

Sketch to Render prompts should include the garment type, style, color, fabrication.

Better (Sketch to Render)

Oversized utility jacket with a boxy silhouette, matte olive nylon fabric, clean seam lines, and minimal brushed gold hardware.

Mistake #6: Forgetting Continuity

  • Inconsistent terms across chained prompts.
  • Example: starting with jacket style and using coat in one of the iterative prompts, causing a longer coat style in the outputs.
    • Sketch to Render utility jacket > Apply Fabric using burgundy crocodile leather swatch > Edit Module prompt to change to double breasted coat instead of jacket, causing a longer coat style.  

Initial Prompt (Sketch to Render)

Oversized utility jacket with a boxy silhouette, matte olive nylon fabric, clean seam lines, and minimal brushed gold hardware.

Incorrect Prompt (Edit AI Prompts)

Initial prompt

Change to red cotton double breasted coat style, keep all styling the same.

Prompt Assist prompt

Transform this olive green utility jacket into a vibrant red cotton double-breasted long coat, meticulously preserving its original relaxed fit, four distinct flap patch pockets, classic shirt collar, and the style of its brass-toned buttons, now integrated into the double-breasted front closure.

Corrected Prompt (Edit AI Prompts)

Initial prompt

Change to red cotton double breasted jacket, keep overall shape and styling the same. 

Prompt Assist prompt

Transform the olive green utility jacket into a vibrant red cotton double-breasted jacket. Meticulously retain its exact relaxed, boxy silhouette, four large flap patch pockets, structured shirt collar, and the style of its brass-colored buttons and stitching, seamlessly integrated into the new double-breasted design.

Mistake #7: Not Reviewing Outputs Before Iterating

  • Rushing into edits without checking the last result.
  • Trash in, trash out - make sure every step looks great.

Pause and check:

  • Is the silhouette correct?
  • Are the proportions usable?
  • Would I want to build on this?

Fix the base output first — then iterate.

Quick Fix Checklist

✓ Focused
One clear goal per prompt. If you’re changing multiple things, split it into steps.

✓ Clear
No vague words like cool, nice, trendy, or pretty without additional context.

✓ Specific
Product + color + material + style (don’t skip the basics).

✓ Guided
Guide the AI with references and the Similarity Slider in Lifestyle Photography, Product Images, Prints and Patterns, and Graphics and Placement Prints.
(try settings around 30–45% for variation without losing the original vibe).

Prompting well isn’t just about having the right idea - it’s about communicating that idea clearly to the AI. Avoiding the most common mistakes will save you time and frustration, helping you turn your vision into usable images faster. 

Once you’ve mastered the fundamentals and dodged the pitfalls, you’ll be ready for the next step: knowing when prompting alone won’t get you across the finish line, and when it’s time to bring in the Edit Module. New article coming soon.

Heading 1

Heading 2

Heading 3

Heading 4

Heading 5
Heading 6

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur.

Caption

Merchandising and design now create together live in meetings—no more weeks of back and forth.”

[[VP of Merchandising] blog-quote-ttl]
[[$30M alternative eCommerce fashion retailer] blog-quote-subttl]

Quote Name Style

Ordered list
  1. Item 1
  2. Item 2
  3. Item 3
Unordered list
  • Item A
  • Item B
  • Item C

Text link

Bold text

Emphasis

Superscript

Subscript

blog-quote-ttl
blog-quote-subttl
April 6, 2026
Mandisa Foster

Introducing Boards: The Future of Collaborative Design Is Here

January 20, 2026
Mandisa Foster

Prompt Chaining in Raspberry AI: How to Build Ideas Step by Step for Better Results

Get better results in Raspberry AI with prompt chaining—iterate with reference images, generate variations, and fine-tune with similarity sliders.

January 23, 2026
Kenisha Liu

The Design Process Of the Winners | Raspberry 2026 Campaign Challenge

Go behind the scenes of the Raspberry New Year 2026 Campaign Challenge and see how the winners built campaign-ready visuals using Raspberry workflows.

Merchandising
Merchandising
Product
Product
Buying
Buying
Design
Design
Ecommerce
Ecommerce
Print generation
Print generation
Industry-insights
Industry-insights
Community
Community
Events
Events
Product tutorials
Product tutorials
AI fashion design workflows
AI fashion design workflows
Product updates
Product updates

Sign up for our newsletter.

We’re constantly enhancing our product. Stay informed with the latest updates.

You’ll get updates and resources delivered straight to your inbox.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Ready to design smarter, faster, and at scale?

Start creating stunning photorealistic designs in just a few clicks.

Book a demo