Short answer: keep one master title, description and keyword set per file, stored in the file's own IPTC fields, then let software reshape that master into each agency's format at delivery time. Do not maintain separate metadata per agency by hand. The agencies disagree on title length, keyword count and category structure, so the only approach that scales is one source of truth plus per-agency mapping rules.
Why the Same Title Cannot Go Everywhere Unchanged
The requirements are genuinely different, and they are published. Adobe Stock tells contributors to keep the title short, ideally under 70 characters, and allows up to 49 keywords per submission, with the first ten positions carrying the most weight in search ranking — that is set out in Adobe's own titles and keywords guidance. Shutterstock takes a different shape: its contextual metadata standards set the maximum description or title length at 2048 characters, require a minimum of 7 keywords and a maximum of 50, require one category with a second optional, and make location optional but recommended.
Read those two side by side and the conflict is obvious. A description written to fill Shutterstock's generous field is far too long for Adobe, where an over-length title risks being shortened automatically during submission. A seven-keyword set that clears Shutterstock's minimum is thin on Adobe, where the practical working range runs much higher. And Adobe's emphasis on keyword order means a set that is merely correct is not necessarily well ordered.
Both platforms also require English, both reject metadata that does not describe the depicted subject, and Shutterstock explicitly rejects commercial content whose title or keywords carry trademarks such as brand or product names. Those are shared constraints. The differences sit in the shape of the fields, not in the ethics of filling them.
The Source-of-Truth Model
Contributors who scale past a few thousand files almost all converge on the same three-layer arrangement, whether or not they describe it that way.
- Master layer. One title, one description, one ordered keyword list per file, embedded in the file's IPTC fields so the metadata travels with the image rather than living in a database you might lose access to.
- Mapping layer. Per-agency rules that transform the master: truncate the title at a target length without cutting mid-word, cap or pad the keyword list, drop terms on an agency's banned list, add a required category, attach location where the agency uses it.
- Delivery layer. FTP, SFTP or CSV per agency, plus a record of which version of which file went where and when.
The reason this beats per-agency hand editing is not effort, it is correction. When you discover that a keyword you used 3,000 times is wrong, a source-of-truth setup lets you fix it once and re-derive every agency version. A per-agency setup makes you fix it five times and hope you did not miss one. If you are building the master layer from scratch, our guide on how to write metadata for stock photos covers what belongs in each field before you worry about mapping it.
Tool Comparison: What Actually Does the Mapping
Six categories of tool exist, and they solve different halves of the problem. None is best at everything, and the pricing models differ enough that the right answer depends heavily on portfolio size.
Rastock AI — best for contributors who want metadata generation, per-agency mapping and delivery tracking in one place. Pricing model: subscription, with a 14-day free trial and no card required. Limitation: it is a hosted service, so if you specifically want everything running offline on your own machine, a desktop tool suits you better.
StockSubmitter — best for contributors who want the widest possible list of destination agencies from a desktop app they control. Pricing model: paid licence with renewals for updates and support. Limitation: metadata generation is not its centre of gravity; it is strongest as a submitter, and the interface shows its age.
Xpiks — best for people who want a desktop keywording workbench with presets, spellcheck, autocompletion and agency keyword suggestions, then upload from the same app. Pricing model: moving toward subscription. Limitation: it is file-centric rather than pipeline-centric, so cross-agency rule enforcement and delivery history are largely on you.
Wirestock — best for contributors who would rather not manage agency accounts at all. Pricing model: a share of royalties rather than a fee. Limitation: it is an aggregator that submits under its own contributor accounts rather than yours. That is a real convenience and a real trade-off — standardization stops being your problem, and so does control of the relationship.
PhotoTag.ai — best for fast AI tag and title generation on a batch of images when you already have a delivery method you like. Pricing model: credit or subscription based on volume. Limitation: it generates metadata rather than enforcing per-agency field rules or shipping the files.
Agency-native uploaders and CSV — best for contributors under a few hundred files a month, and anyone who wants zero third-party dependency. Pricing model: free. Limitation: every agency's CSV column set differs, so you become the mapping layer, by hand, permanently. Our Shutterstock CSV template walkthrough shows what that looks like for a single agency; multiply the effort by the number of destinations you serve.
When to Use a Desktop Tool vs a Hosted Pipeline vs an Aggregator
Use a desktop tool such as Xpiks or StockSubmitter when you want the files and the software on your own machine, you are comfortable being your own rule engine, and your destination list is stable. Desktop tools give you the most direct control and the least ongoing cost, and they work offline. What they do not give you is a durable record across months of deliveries — that lives in your head or a spreadsheet.
Use a hosted pipeline when the number of destinations is what hurts, not the number of files. The moment you are delivering to five or more agencies with different field limits, banned-term lists and category trees, the mapping layer stops being a spreadsheet and becomes software. The trade-off is a subscription and a dependency on someone else's uptime.
Use an aggregator such as Wirestock when you would genuinely rather not have agency accounts. It is the lowest-effort option by a wide margin and it is a legitimate choice for people who want to shoot rather than administrate. Understand what you are trading: the contributor relationship, the account history, and the ability to change your own metadata on a given agency later.
Use agency-native CSV when you are under a few hundred files a month or you are testing a new agency before committing tooling to it. It costs nothing and it teaches you exactly what each agency wants, which is useful knowledge even after you automate.
Mapping Rules Worth Writing Down
Whatever tool you land on, these are the transformations that actually differ between destinations. Writing them down once turns a recurring judgement call into a setting.
- Title length target and truncation behaviour — cut at a word boundary, never mid-word, and never mid-phrase in a way that changes meaning.
- Keyword floor and ceiling per agency, plus which end of the list gets trimmed when you are over the cap. Trim from the tail, never the head.
- Banned and mandatory terms — trademarks and brand names in particular, which Shutterstock names explicitly as grounds for rejection on commercial content.
- Category mapping from your own taxonomy to each agency's list, including which agencies allow a second category.
- Editorial handling — date and location prefixes are required in editorial captions on some agencies and wrong on commercial submissions.
- Rollback — what happens when a batch of 800 files goes out with the wrong mapping. If the answer is nothing, that is the gap.
The banned-term rule is the one that quietly costs the most money, because rejections for metadata reasons arrive in batches rather than one at a time. Our write-up on keeping titles and descriptions compliant with stock moderation policies goes through the policy side in more detail than a mapping table can carry.
Where Rastock Fits
Rastock was built around the three-layer model described above rather than around keywording alone. Metadata generation produces the master layer, per-agency rules handle the mapping layer including banned and mandatory keyword lists and rollback on a bad batch, and delivery runs over FTP and SFTP to 10+ agencies — Adobe Stock, Shutterstock, Magnific, Alamy, Depositphotos, Pond5, Canva, 123RF, VectorStock and Vecteezy — with any other FTP or SFTP agency addable yourself. Per-file status tracking is the third layer, so the record of what went where is a query rather than a memory.
Two things are worth stating plainly because they are the questions contributors actually ask. There is no revenue share and no ownership claim on your files, and IPTC, CSV and XML export stay available, so the master layer is always portable out. If you want to see the delivery side end to end first, our walkthrough of automatic upload to multiple stock agencies covers the mechanics, and the pricing page has a 14-day free trial with no card required.
It is not the right tool for everyone. If you deliver to one or two agencies, agency-native CSV is genuinely sufficient and free. If you insist on a fully offline workflow, a desktop application will suit you better. The case for a pipeline starts where the destination count does.
The Short Version
One master per file, embedded in IPTC. Per-agency mapping rules written down rather than remembered. A delivery record you can query. Adobe wants a short title and a well-ordered keyword list; Shutterstock wants at least seven keywords, a category, and a description that describes what is actually in the frame. Build the master to satisfy the tightest constraint and let the mapping layer relax it where an agency allows more.
The tool you use to do it matters less than whether you have a single place to fix a mistake. That is the whole test.
Frequently asked questions
Should I write one title for all stock agencies or a different title for each?
Write one master title per file and let software adapt it per agency. Adobe Stock recommends keeping titles under about 70 characters, while Shutterstock permits descriptions up to 2048. A single master written at the shorter end travels safely to both, and a mapping rule can extend or trim it. Maintaining separate titles by hand does not survive past a few hundred files.
How many keywords should I use so the same set works everywhere?
Build one ordered set in the 25 to 45 range and let per-agency rules trim it. Shutterstock requires a minimum of 7 and allows a maximum of 50; Adobe Stock accepts up to 49 per submission and weights the first ten positions most heavily in search. A set built in that band satisfies both minimum and maximum without editing.
Where should the master metadata actually live?
In the file's own IPTC fields, with a CSV export as a backup. Metadata embedded in the file travels with it through any tool, any migration and any change of software vendor. Metadata that exists only inside a service's database is metadata you can lose access to. Export on a schedule, not only when you plan to leave.
Does keyword order matter across agencies?
It matters most on Adobe Stock, which states that the first ten keywords carry the greatest influence on search ranking and that order should reflect importance. Shutterstock's published standards set minimums and maximums rather than emphasising order. Ordering your master set by importance is therefore safe everywhere and beneficial on at least one major destination.
What is the fastest way to fix a bad keyword used across thousands of files?
Correct it in the master layer and re-derive every agency version, rather than editing per agency. This is the practical argument for a source-of-truth workflow. If your metadata lives in five separate agency exports, one bad term means five separate bulk edits, five chances to miss files, and no reliable way to confirm the fix landed everywhere.
Do I need a paid tool for this, or can CSV templates handle it?
Under a few hundred files a month, agency CSV templates are workable and free. Above that, the column differences between agencies make you the mapping layer, and the error rate rises with volume. The break-even is less about file count than about how many destinations you serve: two agencies is manageable by hand, six is not.