Adobe Stock, Shutterstock and Alamy all let you rewrite keywords on content that is already live, and none of them send it back through review for it. The work is done in each platform's own portfolio editor, or in a desktop tool such as Xpiks that writes metadata locally before you push it. The real constraint is not permission — it is that rejected files cannot be edited at all, and mistakes made in bulk are expensive to undo.
Most contributors who have been uploading for more than three or four years are carrying keyword debt. It builds up quietly: a set copied from one shoot to the next because the images looked similar, terms that described a 2019 trend nobody searches for now, keywords added in a second language before an agency clarified its policy, and a long tail of files that were tagged in a hurry to hit a submission target.
The question worth answering first is not how to fix it but whether fixing it pays. This piece covers what each agency allows, which tools do the work, and the sequence that keeps a cleanup from turning into a fresh mess.
What each agency lets you change after publication
Adobe Stock
Adobe's guidance on updating content states that content can be edited before submission or after approval and publication, but not during moderation. Titles, keywords, categories and language are all editable through My Portfolio in the Contributor Portal. Changes appear in your portfolio immediately, though updated metadata takes longer to work through the search index.
Shutterstock
Shutterstock's article on updating content puts approved-content editing in the Catalog Manager, reached from Portfolio in your dashboard. Title, keywords and categories can all be changed there without triggering another review. Three restrictions matter for cleanup work: pending content cannot be edited at all, only deleted; rejected content cannot be amended and must be re-uploaded even if the only change is a title; and commercial-versus-editorial designation is fixed once approved, so switching it means deleting and resubmitting.
Alamy
Alamy Image Manager supports batch editing across a series — captions, tags and licence type applied to a whole selection at once. Their image management guidance also confirms that annotations added today go live the next day, once the search engine refreshes, which it does roughly every 24 hours. If your keywords already live in a spreadsheet, Alamy's image management team will apply them for you on request.
One line in that same document reframes the whole exercise. Alamy states plainly that the discoverability bar does not read the quality of your metadata — it simply rises as you enter more searchable information. They go on to say that you do not have to optimise every image, and that relevance and accuracy matter more than filling the quota. A portfolio-wide push to turn every bar green is therefore not the same thing as a portfolio-wide improvement in sales, and treating it as one is the single most common cleanup mistake.
Two Alamy-specific rules are also worth auditing an old library against. Captions should be in English only, because Alamy machine-translates them and your own translations can confuse the system and push images down the results. And personal or sensitive data does not belong in captions and tags, with narrow exceptions for journalism, documented consent, and cases with a clear public benefit. Older archives frequently violate both without anyone noticing.
Tools that handle bulk keyword cleanup
These fall into three groups, and the group matters more than the individual product. Native agency editors change what is on the agency's servers. Desktop tools change what is embedded in your local files. Managed platforms change both but take a position in your royalty stream. None of them is the right answer for every situation.
Native agency editors — Catalog Manager, My Portfolio, Alamy Image Manager
- Best for: small, targeted fixes and anything where you want to see the result immediately on the platform that matters.
- Cost: free, included with the contributor account.
- Main limitation: edits live only on that agency. Fix a keyword set in Catalog Manager and your local files, and every other agency, still carry the old version. Across a full multi-agency portfolio this is where drift starts.
Xpiks
- Best for: contributors who want direct control over embedded metadata. It is a cross-platform desktop application built specifically for microstock work, with keyword suggestion, CSV import and export, and bulk operations such as removing several keywords from many files at once — exactly the operation a cleanup needs.
- Cost: a one-off desktop licence model with free and paid tiers rather than a per-file charge.
- Main limitation: it edits your local files. For images already published, cleaning the local copy does not update what the agency is serving — you still need to re-push or edit on the platform. It is also a tool that rewards being learned properly rather than opened once.
StockSubmitter
- Best for: high-volume contributors who already run it as their submission hub and want metadata and delivery in one desktop application across a wide list of agencies.
- Cost: a low-cost subscription model, historically one of the cheaper options in this category.
- Main limitation: the interface is dense and Windows-centred, and it is built around getting new work out rather than repairing old work already sitting on agency servers.
PhotoTag.ai and similar AI tagging services
- Best for: regenerating keyword sets from scratch when the originals are beyond salvage — you feed the image and get a fresh set rather than trying to repair the old one.
- Cost: credit or per-image pricing, which makes the cost of a large cleanup easy to estimate and easy to underestimate.
- Main limitation: generic tagging describes what is visible, not what a stock buyer searches for, and it has no concept of a given agency's banned or mandatory terms. On a legacy portfolio it also discards conceptual keywords that were genuinely good — the ones tied to use case rather than to objects in frame.
Wirestock and other managed pipelines
- Best for: contributors who would rather hand the whole distribution problem to someone else and accept a smaller share in exchange for not managing it.
- Cost: a commission on royalties rather than a subscription — you pay out of earnings, continuously, rather than up front.
- Main limitation: it is aimed at content going forward, not at repairing an archive you already placed directly with agencies. It also means an ongoing claim on revenue from work you already own outright.
Rastock AI
- Best for: cleanups that span several agencies at once, where the same corrected metadata has to land consistently in 10+ places. It generates titles, descriptions and keywords, applies per-agency rules on the way out — banned terms, mandatory terms, agency-specific limits — and delivers over FTP and SFTP with rollback if a batch goes wrong.
- Cost: a subscription with no revenue share and no ownership claim on your files; IPTC, CSV and XML export stay available so your metadata can leave with you.
- Main limitation: it delivers over FTP, so it cannot reach in and rewrite records that only exist inside an agency's own portfolio editor. For files already published, the practical route is corrected metadata plus the agency's editor, not one instead of the other. And if you sell on a single platform only, a subscription is harder to justify than the free native editor.
When to use which
If the problem is a few hundred files on one agency, use the native editor and stop there. Nothing else will be faster and nothing else is free. Alamy's batch editing in particular handles a same-shoot series in one pass.
If the problem is that your local masters carry bad metadata that keeps propagating into new uploads, fix it at the source with a desktop tool. Xpiks is built for precisely this: strip a bad keyword from a thousand files, then stop re-introducing it.
If the problem is that the same file has three different keyword sets across three agencies, no single-platform tool solves it. That is a distribution problem, and it needs something that treats one master record as the source of truth and pushes it outward with each agency's rules applied.
And if the keywords are not merely messy but wrong — describing a different subject entirely — regeneration beats repair. That is the one case where an AI tagging pass on the image itself is the efficient move, provided you review the output against the agency's rules before it ships.
The order of operations
Start with data rather than instinct. Pull your download and search-term reports and find out which files actually earn and which terms actually convert, because that determines what deserves the effort — we went through the reporting side in which keywords drive your downloads. A cleanup applied evenly across a whole archive is almost always the wrong allocation.
Then take a full export before touching anything. Every serious tool offers CSV export, and the native editors give you no undo. An exported snapshot is the only rollback you will have if a batch operation strips something you wanted to keep.
Fix removals before additions. Wrong and duplicated keywords do active damage — they pull your files into searches they cannot satisfy, and buyers who scroll past learn to scroll past. Adding better terms on top of a set that still contains the bad ones changes very little.
Work in batches small enough to evaluate. Change a few hundred files, wait for the index to catch up — Alamy refreshes roughly daily, the larger agencies take longer — and compare against the files you left alone. Without a control group you cannot tell whether the cleanup helped or whether the season changed. If you are rebuilding sets rather than trimming them, the structural side is covered in how to structure tagged keywords properly.
Finally, fix the intake. A cleanup that is not followed by a change in how new files get keyworded simply schedules the same job again in three years. That is the argument for putting the rules in the pipeline rather than in your head.
Where Rastock fits
Rastock.ai is built around the multi-agency case: one metadata record, per-agency rules applied at delivery, and status tracked per destination so you know what actually landed. Banned and mandatory keyword lists are enforced on the way out rather than checked afterwards, and a batch can be rolled back. The features page covers the delivery side and the pricing page sets out the tiers. There is a 14-day free trial with no card required, which is long enough to run one problem batch through and compare it against what you would have written yourself.
It is not the right tool for everyone. A single-agency portfolio is better served by the free native editor, and a contributor who wants someone else to run distribution entirely will prefer a managed platform even at the cost of a royalty share. The case for a delivery layer starts when the same file has to be correct in several places at once.
Frequently asked questions
Can I edit keywords on stock photos that are already published?
Yes, on all three major platforms. Adobe Stock allows edits to titles, keywords, categories and language after approval through My Portfolio, though not while content is in moderation. Shutterstock allows the same through Catalog Manager without a further review. Alamy Image Manager supports batch edits to captions and tags across a whole selection of images.
Will re-editing keywords hurt my existing search ranking?
Editing metadata does not send approved content back through review, so there is no approval risk. The genuine risk is removing terms that were quietly earning. That is why an export before any bulk operation matters, and why it is worth changing a subset first and comparing it against files you left untouched rather than rewriting an entire archive at once.
Can I fix keywords on rejected files?
Not on Shutterstock. Their documentation is explicit that rejected content cannot be edited or amended and must be re-uploaded and resubmitted, even when the only change is a title. The same generally applies to the commercial versus editorial designation, which is fixed once content is approved and requires deleting and re-uploading the file to change.
Does filling every keyword slot improve discoverability?
Not by itself. Alamy states directly that its discoverability bar does not read the quality of your metadata and simply rises with the volume of searchable information entered. They add that not every image needs the full quota, and that relevance and accuracy matter more. Padding a set to turn the bar green can attract traffic that never converts.
What is the best tool for cleaning up an old stock portfolio?
It depends on where the mess lives. Native agency editors are free and immediate but only fix one platform. Xpiks and similar desktop tools fix embedded metadata in your local masters. Rastock AI suits cleanups spanning several agencies, applying per-agency rules at delivery. AI tagging services are best when sets need regenerating rather than repairing.
Is re-keywording an old portfolio worth the time?
For most contributors, only in part. The work pays where files already receive views but convert poorly, or where keywords are actively wrong. It rarely pays across an entire archive, since a large share of any back catalogue earns almost nothing regardless of metadata. Pull your download reports first and let them decide what gets attention.