Define Style Options

The Visualizer gives you seven ways to define a style. Each serves a different consultation scenario. This guide explains what each option does, what to expect from it, and how to combine options for more precise results.

How Style Definition Works

When you click Generate New Look, the Visualizer sends three things to the AI: the client portrait, a style instruction (the prompt), and optionally a color direction. The style definition section is where you build that instruction.

You are not locked into one method. Options can be layered. The most common combination is a preset baseline with added detail refinements, but any combination is valid.

Option 1: Get AI Recommendations

What it does: Analyzes the uploaded client portrait and returns three style suggestions tailored to that specific person. Each suggestion includes a style name, a rationale explaining why it suits the client, and a generation-ready prompt.

What to expect: Three recommendation cards appear in the Visualizer. Each card shows the style name and the reasoning. Tap Try this Style on any card to apply the prompt and move directly to generation. You can also review the prompt text before generating and add refinements using the Add Details field.

Use this when:

• The client is unsure what they want and needs direction options
• You want a starting point based on the client's actual face rather than a preset library
• You are working with a new client and do not yet know their preferences

Important: AI Recommendations analyze the portrait, not the client's hair type, texture history, or past consultations. Treat the suggestions as informed starting points, not definitive prescriptions.

Option 2: Upload Reference

What it does: Accepts an inspiration photo uploaded by you or provided by the client. The AI analyzes the reference image and extracts a descriptive style prompt based on what it sees.

What to expect: After upload, a generated prompt appears in the style field describing the shape, length, and key characteristics the AI identified in the reference photo. Review this prompt before generating. If the extracted description misses something important (for example, the reference photo had an unusual angle), edit the prompt or add clarifying details using the Add Details field.

Use this when:

• The client arrives with a saved photo from Instagram, Pinterest, or a magazine
• You want to approximate a specific look without manually writing a prompt
• The inspiration style is too specific or unusual to find in the preset library

Note: Reference photo quality affects extraction accuracy. A clear, well-lit, front-facing reference produces a more accurate prompt than a cropped, blurry, or heavily filtered image.

Option 3: Hair Style Presets

What it does: Provides a curated library of common and trending hairstyles organized into Women's and Men's categories. Selecting a preset applies a tested, generation-ready prompt as your style baseline.

What to expect: Presets are displayed as a grid with thumbnail previews. Click any preset to see a full-screen preview and a description of the style. Select it to apply it as your active style direction. A selected preset shows a highlighted border.

You can hide presets you never use. You can also copy any preset as a starting point for a custom style saved to your personal library. See Option 7 for how custom styles work.

Use this when:

• You know the style the client wants and it maps to a recognized cut or trend
• You want a fast, consistent baseline without writing a custom prompt
• You are running back-to-back consultations and need speed

Combining with other options: After selecting a preset, use the Add Details field to refine the result without replacing the base style direction.

Option 4: Hair Color Presets

What it does: Applies a color direction to the generation. Color presets are swatches organized by color family (black and brunette, blonde, red and copper, gray and silver, fashion colors). Each swatch includes a prompt suffix that guides the AI toward realistic color rendering.

What to expect: Two application modes are available:

• Apply to selected style: Combines the active style direction with the color direction. Use this when you want to change both cut and color in the same generation.
• Keep current hairstyle: Applies color while minimizing haircut shape changes. Use this when the client only wants to explore color options on their existing cut.

Use this when:

• The client wants to see a color change on their current style
• You are exploring color options before recommending a chemical service
• You want to combine a new cut with a new color in one preview

Getting cleaner color results: Keep the Add Details field concise when using color presets. Long or conflicting detail prompts can cause the AI to override the color direction. If the haircut shape is changing more than intended, switch to "Keep current hairstyle" mode.

Option 5: Describe a Style from Scratch

What it does: A free-text field where you write a custom style prompt describing exactly what you want the AI to generate.

What to expect: The AI interprets your description and applies it to the client portrait. The more specific and structured the description, the more predictable the result.

Use this when:

• The requested style is too specific, unusual, or niche for the preset library
• You want full control over the generation direction
• You are an experienced user who has developed effective prompt language

Prompt structure that produces better results:

Describe these elements in order when possible:

1. Length (for example: chin-length, shoulder-length, mid-back)
2. Cut shape (for example: blunt ends, layered, tapered at the nape)
3. Texture and movement (for example: straight, wavy, voluminous at the crown)
4. Fringe (for example: no fringe, curtain bangs, blunt fringe)
5. Part (for example: center part, deep side part)
6. Color, if not using a color preset (for example: warm brunette with face-framing highlights)

Example prompt: "Shoulder-length lob with soft, blended layers, slight wave, curtain bangs parted in the center, warm chestnut brown."

Avoid vague direction words like "natural," "beautiful," or "modern" without pairing them with structural descriptors. The AI responds to shape and texture language more reliably than aesthetic judgments.

Option 6: Add Details to Refine This Style

What it does: An additional text field that appears when a preset or reference-driven base style is active. It lets you modify the base direction without replacing it entirely.

What to expect: Your refinement text is appended to the base style prompt before generation. The AI applies the base style first and then attempts to incorporate your refinements.

Use this when:

• A preset is close but not quite right
• The client wants a small adjustment to a reference-extracted style
• You want to test variations on the same base style across multiple generations

Effective refinements:

• "Softer fringe density, more face-framing"
• "Slightly shorter at the back, keep the length at the front"
• "Add subtle warm highlights through the mid-lengths"
• "Reduce volume at the crown"

Avoid:

• Contradicting the base style entirely (use Describe from Scratch instead)
• Writing a second full prompt in the refinement field
• Long, multi-sentence refinements that conflict with each other

Option 7: Custom Styles (My Styles)

What it does: Your personal style library. Custom styles are presets you create yourself -- saved combinations of a style name, a prompt, and a thumbnail image. They appear in the My Styles tab alongside the Women's and Men's preset libraries.

Custom styles belong to your account, not the salon. They are available on any device you log in from and move with you if you change workplaces.

How many custom styles you can save depends on your plan:

• Free: 0
• Starter: 12
• Pro: Unlimited
• Salon: Unlimited

Two ways to create a custom style:

From a generation result (recommended): After a generation you are happy with, click Save as Custom Preset below the result image. A modal opens prefilled with the generated image as the thumbnail, the prompt that produced the result, and a suggested style name. Review and edit any field before saving. This approach saves a recipe you have already proven works.

Manually in the Visualizer: In the Define Style section, click Create Preset Manually. Enter a style name, write the prompt from scratch, and upload a reference image to use as the thumbnail. This is the faster path for experienced users who know exactly what prompt they want to save without generating first.

Editing and managing custom styles: Custom styles can be edited or deleted at any time from the My Styles tab. You can also update the thumbnail image after saving -- for example, if you later generate a better result that better represents the style.

Use custom styles when:

• You have a go-to look you recommend frequently and want one-tap access to it
• A client returns for the same service and you want to reproduce a previous result consistently
• You have developed a prompt that reliably produces a specific outcome and want to lock it in

Combining Options: Common Patterns

What Comes Next

• Your First Consultation -- end-to-end walkthrough including generation and saving
• Best Practices -- photo quality guidance and prompt strategies for consistent results
• Troubleshooting -- fixing unexpected generation results