Every stock agency asks for the same information in a different shape, which is why one metadata set rarely survives contact with more than one destination. Adobe Stock wants a title of 70 characters or fewer with no commas in it. Shutterstock wants a description that reads like a sentence. Getty wants files at 3 megapixels while Shutterstock rejects anything under 4. Mapping, not writing, is the real work.
The field rules that actually differ
Before choosing a tool it helps to see how far apart the specifications sit. These are taken from the agencies' own contributor documentation, which is the only version worth trusting, since third-party guides go stale quickly and the numbers move.
Adobe Stock
- Filename: 30 characters maximum, including the extension.
- Title: 70 characters maximum, plain text, no commas or special characters.
- Keywords: 50 maximum, comma-separated, ordered by importance.
- Category: a number, not a word, taken from Adobe's own category list.
- Releases: the exact release filename, JPEG, also capped at 30 characters.
- CSV itself: .csv only, no .xls or .xlsx, maximum 5,000 rows, maximum 1 MB, UTF-8, and the header row must match Adobe's template exactly with no renaming or translation.
Shutterstock
- Description, not title: it should read as a sentence or phrase answering who, what, when, where and why, and explicitly not as a list of words.
- Keywords: between 7 and 50. The floor matters, because Adobe has no minimum and a six-keyword file that passes there will fail here.
- Forbidden in both fields: links, specific camera information, and trademarks, except where relevant to editorial content.
- Files: JPEG or unlayered TIFF, sRGB, minimum 4 megapixels. EPS must be Illustrator 8 or 10 compatible and between 4 and 25 megapixels measured on the bounding box, not the artboard.
- Upload ceilings differ by route: through a browser it is 50 MB for photos and 100 MB for EPS, while over FTPS you can send JPEGs up to 50 MB, EPS up to 100 MB and TIFFs up to 4 GB.
Getty Images and iStock
- Files: JPEG, RGB, minimum 3 megapixels, maximum 256 megapixels, highest JPEG quality, no upsizing.
- Generative AI content is not accepted in any form, and no metadata flag changes that.
- Upload access is granted only after a portfolio review, so the mapping question does not arise until you are in.
Read together, these produce awkward collisions. A 3.5 megapixel file clears Getty and fails Shutterstock. A comma-rich Shutterstock description breaks an Adobe CSV. A file named with a long descriptive string passes Shutterstock and is truncated out of Adobe's 30-character limit. The per-agency detail is covered separately in Adobe Stock's metadata requirements and in Getty's contributor requirements.
The tools contributors use for this, and what each is actually good at
There are broadly three approaches: do the mapping yourself in a spreadsheet, use a desktop application that holds per-agency profiles, or use a hosted service that maps and delivers for you. None of them is best in every case.
Agency CSV templates plus a spreadsheet
Best for: small, occasional batches and for anyone who wants to understand the rules before automating them. Cost model: free. Main limitation: it does not scale and it fails silently. Adobe's 1 MB and 5,000-row ceilings arrive without warning, and a single renamed header rejects the whole file. Every agency needs its own sheet, which means the same edit has to be made in several places.
Xpiks
Best for: contributors who want a genuinely cross-platform desktop tool and prefer their files and metadata to stay local. Cost model: free. Strengths worth stating plainly: it handles photos, vectors and video, has keyword presets, spellcheck and autocompletion, and uploads to multiple agencies. Main limitation: it is a desktop application, so the mapping runs on your machine and your batch pipeline is tied to that machine being on.
StockSubmitter
Best for: contributors who need the widest agency coverage and do not mind an older interface. It supports a large number of microstock destinations, which is its real strength and the reason long-time contributors still recommend it. Cost model: has a free tier. Main limitation: Windows only, so Mac and Linux users are out, and the interface shows its age.
PhotoTag.ai
Best for: generating keywords and titles quickly when the destination is a single agency or your own site. Cost model: credit-based. Main limitation: it is a tagging tool rather than a delivery pipeline, so the mapping to each agency's field limits and the upload itself remain yours to handle.
Wirestock
Best for: contributors who want the whole submission problem taken off their hands and are comfortable with a managed relationship. Cost model: takes a share of royalties. Main limitation: that share is the trade, and you have less direct control over how each agency sees your metadata.
Rastock AI
Best for: contributors who want one metadata set mapped and delivered to several agencies without a royalty share. It generates titles, descriptions and keywords, applies your own banned and mandatory keyword rules so a policy breach is caught before it ships, and delivers over FTP or SFTP to 10+ agencies with per-file status. Cost model: flat subscription, with a 14-day trial that does not ask for a card. Main limitation: it is built around agencies that accept FTP or SFTP, so a destination that only takes manual web upload still needs a manual step. The pipeline is described here.
When to use a spreadsheet, a desktop tool, or a pipeline
Use a spreadsheet when you are sending a few hundred files a month to one or two agencies. The overhead of learning a tool is not repaid at that volume, and doing the mapping by hand once teaches you which rules actually bite.
Use a desktop tool such as Xpiks or StockSubmitter when volume has grown but you want your originals and metadata to stay on your own hardware, or when you need an agency that only those tools support. The trade is that your throughput is bounded by one machine.
Use a hosted pipeline when the same set goes to four or more agencies regularly and re-typing has become the bottleneck. Within that group, the real decision is whether you are willing to give up a share of royalties. A managed service such as Wirestock takes one; a subscription tool does not. Neither is wrong, but the arithmetic changes sharply once your monthly earnings pass the cost of a subscription.
How to build the mapping so it survives a rule change
Whichever route you pick, the structure that lasts is the same. Keep one canonical record per file, written to the loosest common denominator: a descriptive sentence, a short label, and an ordered keyword list of 25 to 50 terms. Then derive each agency's version from it rather than editing agency copies directly.
Derive the Adobe title by truncating the canonical label to 70 characters with commas stripped. Derive the Shutterstock description from the canonical sentence unchanged. Take the first 50 keywords for both, and check the count is at least 7 before anything goes to Shutterstock. Store releases under short filenames from the start, because renaming them later to fit a 30-character limit breaks the reference.
The reason to derive rather than duplicate is that agencies revise these rules. When a limit moves, you change one transformation instead of auditing thousands of rows across five spreadsheets. The same principle underlies standardising titles and descriptions across agencies, and it is what separates a workflow that scales from one that merely works today.
The mistakes that cost the most time
Translating a CSV header is the classic one. Contributors working in a non-English version of Excel open Adobe's template, the software localises the column names, and the entire upload is rejected with no useful explanation. The related trap is saving as .xlsx out of habit when only .csv is accepted, or saving without UTF-8 encoding so that every accented character arrives corrupted.
The second is treating keywords as a single global list. A term that is precise at one agency can be spam at another, and camera data, trademarks and file names do not belong in the field at any of them. The third is discovering the megapixel floor after processing a batch rather than before, which turns a five-minute export setting into a full reprocess.
None of this is difficult in isolation. It is difficult at volume, across five destinations, on a Sunday evening. If you have reached the point where mapping is the bottleneck rather than the shooting, it is worth looking at what a delivery pipeline costs against the hours it replaces, and being honest about which of the two you are actually short of.
Frequently asked questions
Can I use the same title and keywords for every stock agency?
Not without adjustment. Adobe Stock caps titles at 70 characters and disallows commas, while Shutterstock expects a description written as a full sentence, which usually contains commas and runs longer. Keywords are closer: both cap at 50, but Shutterstock also sets a minimum of seven. The practical approach is to keep one canonical record and derive each agency's version from it.
What are Adobe Stock's CSV file limits?
Adobe accepts .csv only, not .xls or .xlsx, with a maximum of 5,000 rows and a maximum file size of 1 MB. The file must be UTF-8 encoded, which matters for contributors using non-US versions of Excel, and the header row has to match Adobe's official template exactly. Renaming or translating a column header will cause the upload to fail.
Why does a file accepted by Getty get rejected by Shutterstock?
Most often it is resolution. Getty accepts files from 3 megapixels upward, while Shutterstock requires at least 4, so anything in that gap passes one and fails the other. Colour space is another common cause, since Shutterstock expects sRGB and converts anything else, which can shift your colours. Generative AI content is a third: Adobe accepts it when labelled, Getty does not accept it at all.
Do I need a paid tool to map metadata across agencies?
No. At low volume a spreadsheet built from each agency's own template works, and free desktop tools such as Xpiks handle photos, vectors and video with per-agency profiles. Paid tools become worth considering when the same set goes to four or more destinations regularly and the manual mapping step is what limits how much you can upload.
Should I choose a royalty-share service or a subscription tool?
It depends on volume and on how much control you want. A managed service such as Wirestock removes almost all of the work in exchange for a share of what your files earn, which suits contributors who would otherwise not submit at all. A flat subscription costs the same whether you earn nothing or a great deal, so it favours contributors with steady output. Work out both against your current monthly earnings rather than in the abstract.
How many keywords should I actually use?
Between 25 and 50 is a reasonable working range for most files, ordered by relevance rather than alphabetically, since several agencies weight the earliest terms most heavily. Fifty is the ceiling at both Adobe Stock and Shutterstock, and seven is Shutterstock's floor. Padding a list to reach the maximum with loosely related terms is counted as spam and is a common reason for metadata rejections.