How to Use Midjourney for Product and Campaign Visuals

Midjourney can produce product and campaign visuals that are usable for social media, pitch decks, mockups, and even direct marketing assets. The challenge is not generating an image. The challenge is generating an image that looks intentional, matches brand requirements, and works at the right dimensions for its destination.

This guide covers the techniques that matter for commercial visual work: controlling composition with aspect ratios, using --no text to keep type out of the frame (so you can add it yourself later), leveraging references and personalization for brand consistency, and structuring prompts that produce predictable, usable results.

Why product prompts are different from creative prompts

Creative prompts aim for surprise. Product prompts aim for control. When you generate a campaign visual, you typically need a specific composition, a specific mood, a specific subject placement, and a specific aspect ratio. You also need the image to be free of AI-generated text because you will add real typography in your design tool.

The biggest mindset shift: you are not asking Midjourney to be creative. You are asking it to be precise. Your prompt is a creative brief, not an inspiration request.

Aspect ratios for campaign destinations

Every marketing asset has a target dimension. Set your aspect ratio at generation time, not in post. Cropping after generation loses composition intent.

• --ar 1:1 for Instagram feed, product thumbnails, and square ad units
• --ar 4:5 for Instagram portrait posts and Pinterest pins
• --ar 9:16 for Stories, Reels, TikTok, and vertical ad placements
• --ar 16:9 for YouTube thumbnails, blog headers, and landscape banners
• --ar 3:2 for website hero sections and email headers
• --ar 2:3 for tall format posters and event graphics

a minimalist perfume bottle on white marble, studio lighting, soft shadows, negative space on the left for text placement --ar 4:5 --no text words letters

Removing AI-generated text with --no text

Midjourney will sometimes add decorative text, labels, or letterforms into images, especially when the prompt describes anything commercial. This looks like a design artifact and makes the image unusable for real campaigns where you need to overlay actual typography. The --no parameter followed by text-related terms prevents this.

a luxury skincare product in a botanical setting, warm natural light, editorial photography style --ar 3:2 --no text words letters typography labels

flat lay product photography, headphones on a concrete desk, minimal props, top-down angle --ar 1:1 --no text words writing logos

Always add --no text words letters to any product or campaign prompt. This is the single most important parameter for commercial work. Without it, you will regenerate images repeatedly just to get a clean frame.

Structuring product prompts

A strong product prompt has five components in order: subject, environment, lighting, style reference, and composition notes. Each component narrows the output toward something usable rather than generic.

The five-part product prompt formula

• Subject: what the product is, specific details, material, color
• Environment: surface, background, context (studio, lifestyle, flat lay)
• Lighting: direction, quality, temperature (warm, cool, dramatic, soft)
• Style: photography style or reference (editorial, commercial, documentary)
• Composition: camera angle, framing, negative space, text placement areas

a matte black water bottle on a granite countertop, warm morning light from the left, lifestyle photography, eye-level angle, negative space above for headline --ar 16:9 --no text words letters

Example: social media campaign set

premium wireless earbuds in an open charging case, placed on a wooden desk, soft window light, shallow depth of field, overhead angle, minimal background --ar 1:1 --no text words logos

the same wireless earbuds being worn by a person walking in an urban environment, natural daylight, street photography style, medium shot --ar 9:16 --no text words

Using references for brand consistency

For campaign work that needs to match existing brand photography, you can use Midjourney's reference system to anchor the visual style. Upload a reference image that represents your target look, and Midjourney will use it to inform lighting, color grading, composition patterns, and overall mood.

This works best when your reference image has strong, identifiable characteristics: a specific color palette, a consistent lighting setup, or a distinctive editing style. Generic references produce generic guidance.

[brand_reference_image_url] a ceramic coffee mug on a linen surface, soft directional light, earth tones, editorial product photography --ar 3:2 --no text words

Personalization for repeat campaign work

If you produce campaign visuals regularly and have trained Midjourney's personalization profile through consistent voting and usage, the --p parameter applies your aesthetic preferences globally. This is useful for agencies or brands that want every generation to lean toward a known visual identity without uploading references each time.

a hand holding a smartphone displaying a gradient wallpaper, studio background, soft fill light --ar 9:16 --p --no text words labels ui

Campaign visual recipes

Hero banner with text space

a sunlit kitchen with fresh vegetables on the counter, warm tones, lifestyle photography, wide shot with empty space on the right third for overlay text --ar 16:9 --no text words letters

Instagram carousel product set

a set of three glass skincare bottles on a marble shelf, backlit, soft caustics, pastel gradient background, centered composition --ar 1:1 --no text words logos labels

Event announcement background

abstract geometric shapes in deep navy and gold, subtle texture, dark background, minimal, elegant atmosphere --ar 9:16 --no text words

Common mistakes in product prompts

• Forgetting --no text. This is the single most common issue with commercial Midjourney work
• Using --ar after the fact. Set dimensions at generation time for proper composition
• Describing too many products in one frame. Midjourney handles single-subject shots better
• Skipping lighting direction. Without it, results are flat and generic
• Not specifying negative space for text. If you need overlay room, say where in the prompt
• Using vague style terms like 'professional' or 'modern' without anchoring to a photography genre

Iterating toward final assets

Commercial work rarely lands on the first generation. Plan for 3 to 5 rounds of iteration. On each round, identify what is wrong (lighting too flat, subject too small, background too busy) and adjust one element at a time. Changing multiple prompt elements simultaneously makes it impossible to track what improved the output.

• Round 1: establish subject and composition
• Round 2: refine lighting and mood
• Round 3: lock aspect ratio and negative space
• Round 4: apply references or personalization for brand match
• Round 5: generate variations for A/B testing

Batch generation for campaign sets

Most campaigns need multiple assets in different formats. Rather than prompting from scratch for each dimension, start with your strongest composition at its native aspect ratio. Once you have one image you are happy with, use that result as a reference for other aspect ratios. This maintains the visual identity while adapting the framing for different placements.

Original (feed post): premium headphones on a slate surface, warm backlight, editorial style --ar 1:1 --no text words letters
Story variant: [feed_image_url] premium headphones, same lighting, vertical crop with space at top --ar 9:16 --no text words letters
Banner variant: [feed_image_url] premium headphones, same lighting, wide crop with left-side negative space --ar 16:9 --no text words letters

This approach lets you build a complete campaign set from a single winning generation rather than starting from zero for each placement size.

Next steps

Related reading: /blog/how-to-use-midjourney-image-prompts-and-references covers reference techniques in depth. /blog/midjourney-parameters-and-aspect-ratio-explained provides a full parameter reference for aspect ratios and other modifiers. For scaling this into a repeatable system, see /blog/how-to-build-consistent-visual-systems-with-midjourney.