How to Keyword Stock Photos for Shutterstock (With CSV Export)
Keywording is often the slowest part of submitting stock photos. A strong shoot can turn into an evening of writing descriptions, repeating keywords, checking categories, and copying everything into the Shutterstock Contributor portal.
The faster approach is to treat metadata as a batch workflow. First select the images worth submitting. Then create accurate descriptions and keywords. Finally, place that metadata in a CSV that Shutterstock can match to the image filenames.
This guide shows the complete process.
## What Shutterstock needs in a metadata CSV
Shutterstock accepts metadata for image and video submissions as a CSV file. According to its Contributor help page, every row represents one file. The required columns are:
- Filename
- Description
- Keywords
- Categories
The filename must match the uploaded asset. Shutterstock says one or two categories can be included, separated by commas. Optional columns can also identify illustration, mature content, or editorial content.
The column names and formatting matter. Shutterstock warns that a CSV that does not match its required structure can be rejected. Upload the photos or videos first, then use the CSV button on the Submit page to add the metadata file. Do not upload the CSV with the media files through FTPS or the regular web upload area.
## A practical keyword strategy
A long list is not automatically a useful list. Every keyword should describe something visible, place the image in a useful context, or express a concept a buyer could reasonably search for.
Build the list in layers:
1. Main subject. Start with the clearest nouns: woman, hiker, mountain, laptop, street, dog.
2. Action. Add what is happening: walking, working, cooking, running, smiling.
3. Setting. Describe the environment: office, outdoors, city, forest, beach, home.
4. Visual details. Include relevant colors, weather, time of day, viewpoint, composition, and copy space.
5. Concepts. Add ideas supported by the image: teamwork, freedom, travel, sustainability, remote work.
6. Location. Use an exact place only when you can identify it confidently and it helps a buyer find the image.
Avoid keywords that are only loosely related. If a buyer searching that term would feel misled by the result, remove it. Also remove duplicates, spelling variants that add no value, camera specifications, and names or brands that should not appear in commercial metadata.
## How to write a useful description
Write one clear sentence that identifies the subject, action, and setting. Add a distinctive detail only when it helps separate the image from similar results.
A simple formula is:
[main subject] + [action or condition] + [setting] + [useful visual detail]
Get StockFlow’s stories in your inbox
Join Medium for free to get updates from this writer.
Example:
Woman working on laptop at a cafe table beside a large window with copy space
This is more useful than “Business lifestyle concept” because it tells the buyer what is actually visible. It is also better than a camera-heavy description that mentions the lens, exposure, or editing style.
## Keyword the image, not the whole shoot
A common batch mistake is applying the same keyword list to every frame. Related photos can share a base set, but each image still needs a final review.
Check what changes from frame to frame:
- Is the person smiling, looking away, or using a device?
- Is the shot wide, close, vertical, or horizontal?
- Is there copy space, and where is it?
- Does a landmark appear clearly in one frame but not another?
- Is the best image commercial, editorial, or unsuitable to submit?
This is also why culling belongs before keywording. Writing metadata for ten near-identical frames wastes time and can lead to a weak submission. Choose the strongest frame first.
## How to build the CSV
Create one row per file and keep the filenames exact, including the extension. Put the description, keyword list, and category value in the matching columns. If a field contains commas, your spreadsheet or CSV tool must quote and escape the field correctly.
Before upload, check:
- Every CSV filename matches an uploaded file.
- Required headers use Shutterstock’s exact names.
- Keywords are relevant and separated correctly.
- Categories use Shutterstock’s accepted values.
- Optional illustration, mature, and editorial fields are included only when needed.
- The file was saved as CSV, not XLSX.
## A faster folder-to-CSV workflow
StockFlow combines the steps that usually happen in separate tools. Open a folder in the browser, review a clean image grid, and let the AI score each photo from 1 to 5. Similar shots are flagged so you can keep the strongest frame and skip the rest. For approved images, StockFlow generates a descriptive title, 50 targeted keywords, and categories. You can edit the result before exporting a Shutterstock-formatted CSV.
The point is not to remove your judgment. It is to start from a complete batch draft instead of a blank field for every image.
StockFlow includes 50 images per month free, with no credit card required. Try it on one finished shoot and compare the time from folder selection to CSV export.
## Sources
Shutterstock Contributor: How do I include existing metadata with my content submission?
https://submit.shutterstock.com/help/en/articles/10617486-how-do-i-include-existing-metadata-with-my-content-submission
StockFlow product page
https://stock-flow.io/
