Best Practices
Tips and tricks to get the most out of Adogen and maximize your Adobe Stock sales.
---
title: "Best Practices"
description: "Tips and tricks to get the most out of Adogen and maximize your Adobe Stock sales."
category: "Getting Started"
lastUpdated: "Fri Jun 05 2026 12:39:00 GMT+0000 (Coordinated Universal Time)"
---
### Before You Upload
- **Use well-lit images** with clear subjects. AI models perform best with good lighting and composition.
- **Avoid cluttered backgrounds.** If the main subject is unclear, the AI may misidentify it.
- **Crop tightly.** Focus on the main subject to give the AI the best chance of accurate analysis.
### Crafting Better Metadata
- **Experiment with multiple engines.** Groq, Mistral, and OpenRouter may give different results for the same image. Use the engine that produces the most relevant keywords for your style.
- **Review and refine.** AI-generated metadata is a starting point. Review the title and keywords before uploading — remove irrelevant keywords and add niche terms the AI might have missed.
- **Use custom prompts.** In Settings, add a custom prompt like "Focus on colors and textures" or "Use formal, professional language" to steer the AI toward your preferred style.
### Batch Processing Tips
- **Group similar images together.** Processing a batch of similar images (e.g., all from the same shoot) helps the AI maintain consistency in keywords and categories.
- **Set an appropriate delay.** If you're processing 50+ images, set the delay to 3–5 seconds to avoid hitting rate limits.
- **Review in batches.** After each batch, scan the results for any obvious errors before exporting.
### Keywords Strategy
- **Aim for 40–50 keywords.** Adobe Stock rewards comprehensive keyword lists. Include synonyms, related concepts, and variations.
- **Order matters.** The AI places the most important keywords first — this is what buyers see in search results.
- **Think like a buyer.** What words would someone type to find your image? Include those.
### Title Strategy
- **Be factual, not creative.** Adobe Stock titles should describe what's in the image, not how it makes you feel.
- **Avoid subjective words.** Don't use "beautiful," "amazing," "stunning" — stick to objective descriptions.
- **Include key details.** Colors, setting, number of subjects, and perspective all belong in a good title.
### Category Selection
- **Trust the AI, but verify.** The AI's category choice is usually correct, but double-check if your image could fit multiple categories.
- **People images go to Category 13.** If your image has recognizable people, make sure it's categorized correctly.
- **Lifestyle vs. Business.** Ask yourself: "Would a buyer looking for business content find this?" If yes, use Category 3.
### Workflow Integration
- **Export CSV regularly.** Don't let too many results pile up — export after each batch to keep things organized.
- **Use History as a reference.** If you shoot similar content regularly, check your History for past metadata you can reuse or adapt.
- **Save engine presets.** If you switch between engines for different types of content, save each configuration as a preset for quick access.
Before You Upload
- Use well-lit images with clear subjects. AI models perform best with good lighting and composition.
- Avoid cluttered backgrounds. If the main subject is unclear, the AI may misidentify it.
- Crop tightly. Focus on the main subject to give the AI the best chance of accurate analysis.
Crafting Better Metadata
- Experiment with multiple engines. Groq, Mistral, and OpenRouter may give different results for the same image. Use the engine that produces the most relevant keywords for your style.
- Review and refine. AI-generated metadata is a starting point. Review the title and keywords before uploading — remove irrelevant keywords and add niche terms the AI might have missed.
- Use custom prompts. In Settings, add a custom prompt like “Focus on colors and textures” or “Use formal, professional language” to steer the AI toward your preferred style.
Batch Processing Tips
- Group similar images together. Processing a batch of similar images (e.g., all from the same shoot) helps the AI maintain consistency in keywords and categories.
- Set an appropriate delay. If you’re processing 50+ images, set the delay to 3–5 seconds to avoid hitting rate limits.
- Review in batches. After each batch, scan the results for any obvious errors before exporting.
Keywords Strategy
- Aim for 40–50 keywords. Adobe Stock rewards comprehensive keyword lists. Include synonyms, related concepts, and variations.
- Order matters. The AI places the most important keywords first — this is what buyers see in search results.
- Think like a buyer. What words would someone type to find your image? Include those.
Title Strategy
- Be factual, not creative. Adobe Stock titles should describe what’s in the image, not how it makes you feel.
- Avoid subjective words. Don’t use “beautiful,” “amazing,” “stunning” — stick to objective descriptions.
- Include key details. Colors, setting, number of subjects, and perspective all belong in a good title.
Category Selection
- Trust the AI, but verify. The AI’s category choice is usually correct, but double-check if your image could fit multiple categories.
- People images go to Category 13. If your image has recognizable people, make sure it’s categorized correctly.
- Lifestyle vs. Business. Ask yourself: “Would a buyer looking for business content find this?” If yes, use Category 3.
Workflow Integration
- Export CSV regularly. Don’t let too many results pile up — export after each batch to keep things organized.
- Use History as a reference. If you shoot similar content regularly, check your History for past metadata you can reuse or adapt.
- Save engine presets. If you switch between engines for different types of content, save each configuration as a preset for quick access.