Originally published at clipmeta.app/blog/stock-footage-metadata-guide
If your footage is not showing up in searches, the problem is almost always the same: your metadata is weak. Stock footage metadata is the set of information that describes what is in your clips. Platforms like Blackbox, Shutterstock, Adobe Stock, and Pond5 rely entirely on that data to serve your footage to buyers. No metadata means no discoverability. And poor metadata is almost as bad.
This guide covers everything you need to know to get it right. What Is Stock Footage Metadata? Metadata is structured information attached to your video files. When a buyer searches “sunrise timelapse mountain” on a stock platform, the platform does not watch your video. It reads your metadata and decides whether your clip is a match. Think of metadata as your clip’s listing. A well-written listing brings buyers. A thin listing gets buried. The 5 Core Components
Title
The title is the most important field. It tells the platform what the clip is about and is heavily weighted in search rankings. A good title is specific and descriptive. It names the subject, the action, and when possible, the location. Bad title: “City aerial”
Good title: “Aerial drone shot of downtown Miami skyline at golden hour, Florida” Keep titles under 200 characters. Lead with the most important information. Do not keyword-stuff or add a string of unrelated terms.
Description
The description gives you space to expand on the title. Write 2 to 4 sentences describing what is in the frame, the mood, the camera movement, and any relevant context. Descriptions help with long-tail search queries. Buyers searching for very specific shots often use natural language, and your description is where those phrases live.
Keywords
Learn about Medium’s values
Keywords are the backbone of stock footage discoverability. Most platforms allow 25 to 50 keywords per clip. Use all of them. Good keywords cover the main subject, the action, the setting, the mood, technical details like aerial or slow motion, and conceptual terms like freedom or teamwork. Avoid repeating the same word in different forms unless both forms are genuinely useful.
Categories
Every platform has its own category taxonomy. Choosing the right categories helps the platform serve your footage to buyers browsing by genre rather than search. Map your clip to the most specific category available.
Editorial Status
Editorial footage depicts real identifiable people or private property without a model or property release. Always mark your footage correctly. Mislabeling editorial footage as commercial can get your clip rejected or pulled. Platform-Specific Requirements Blackbox.global uses a CSV upload system. Required columns are filename, title, description, keywords (semicolon-separated), and category. Blackbox has a specific category list so review it before selecting. Shutterstock allows up to 50 keywords and prefers comma-separated values in their CSV. Titles should be under 200 characters. They have a strict review process and will reject clips with generic or misleading metadata. Adobe Stock allows up to 49 keywords and uses a tab-separated CSV format. They pay close attention to keyword relevance, so avoid padding with unrelated terms. Pond5 allows longer descriptions and has a generous keyword limit. They also allow you to set your own pricing, so a detailed description can support a premium price point.
Common Mistakes That Hurt Your Rankings Too few keywords. Using 10 keywords when the platform allows 50 is leaving money on the table. Fill every slot with relevant terms. Generic titles. “Beautiful landscape” tells a platform nothing. “Aerial view of autumn forest with river winding through valley at sunset” tells it everything. Wrong categories. Putting a food clip in Lifestyle instead of Food and Beverage means buyers browsing food footage will never see it. Keyword repetition. Entering “aerial, aerial view, aerial shot, aerial footage” is wasting four slots. Pick the most searched version and move on. Copy-pasting the same keywords for every clip. A clip of a dog running on a beach and a clip of a city at night do not share the same keywords. How AI Metadata Tools Change the Equation The biggest barrier to good metadata has always been time. Writing a good title, a solid description, and 40+ relevant keywords for a single clip takes 10 to 15 minutes done properly. Scale that to 500 clips and you are looking at 80+ hours of work. AI metadata tools like ClipMeta solve this by reading the actual footage rather than guessing. The system pulls frames from each clip, runs them through a vision model, and generates titles, descriptions, and keywords based on what is actually visible in the video.
The result is metadata that reflects the specific content of each clip, not just generic terms that might apply to anything. You review and edit before exporting. It cuts the time per clip from 10 minutes to under a minute. Best Practices for Higher Rankings Be specific first, broad second. Start your keyword list with the most specific terms that describe your clip, then work outward to broader categories. Think like a buyer. What would someone type into a search bar to find this exact clip? Update old clips. Going back and improving thin metadata can revive those clips in search rankings. Match your keywords to your title. The most important keywords should appear in both. Include technical terms. Buyers often search by shot type. “4K aerial drone footage slow motion” is a real search query. Frequently Asked Questions How many keywords should I use for stock footage?
Use the maximum allowed by each platform. Most platforms allow between 25 and 50 keywords. Using fewer limits your discoverability without any benefit. What makes a good stock footage title? A good title is specific, descriptive, and leads with the most important information. Include the subject, action, and location when possible. Does better metadata actually increase sales? Yes. Discoverability is the main driver of stock footage sales. A clip that does not appear in search results effectively does not exist for buyers. How long does it take to write metadata for 100 clips? Manually, expect 10 to 15 hours. With an AI tool like ClipMeta, you can process a batch of 100 clips in roughly 30 minutes including review time. Ready to stop writing metadata by hand? ClipMeta uses AI to generate accurate, platform-ready metadata from your actual footage. Start free at clipmeta.app — 3 clips per day, no credit card required. Use code DEVTO50 for 50% off your first 3 months.
Top comments (0)