Why Stock Photos and Video Get Rejected
A rejection notice tells you a reason in two or three words and nothing else. "Similar Content." "Image Quality." "Commercial Requirements." What it does not tell you is which of a dozen possible causes applied, whether you can fix it, or whether fixing it and resubmitting will get the file banned. This is what each of the major rejection reasons actually means, taken from the agencies' own moderation documentation — organised by reason rather than by agency, because the same reasons recur everywhere with different names.
Similar content: the biggest single cause, and the one people argue with
Adobe calls similar content one of the top reasons it refuses submissions. Shutterstock maintains a dedicated rejection reason for it. And it is the rejection contributors resist most, because from the photographer's chair the frames genuinely look different.
The agencies publish specific thresholds, which is more than most contributors realise.
Adobe: three variations of one scenario
Adobe's guidance is to curate down to three significant versions of a single motif unless the complexity of the setup genuinely calls for more. It explicitly names the failure case: sequential shots from the same moment or location with minimal meaningful difference. Its recommended fix is to vary framing, subject interaction, environment or story — not exposure and crop.
Two useful exceptions sit inside that rule. Adobe says contributors should feel free to submit more than three variations of video, because editors often need several similar clips to cut between. And for AI-generated content it tightens the other way — a maximum of three iterations, with excessive similar submissions treated as content spam.
Shutterstock: the specific list, including one that surprises people
Shutterstock names the patterns it rejects as Similar Content:
- Large sets of visually similar abstract backgrounds and textures.
- The same icon or design with only a different stroke width or treatment.
- The same subject with minimal or no change to framing or pose — including minor adjustments to composition, cropping or aspect ratio.
- The same subject matter shot at different clip lengths or with different camera movements — tilts, pans and pushes.
That last one catches experienced videographers regularly. Covering a scene properly means shooting a static, a slow pan and a push in — which is good production practice and, submitted as three clips of the same subject, is exactly the pattern Shutterstock names. The fix is not to stop covering the scene, it is to change the subject or the setup between coverage passes rather than only the movement.
Getty: two of three must differ
Getty applies the clearest arithmetic rule of the three, in its brief guidance: two of Models, Location and Wardrobe must differ for a shot not to count as a similar. Same model, same location, new outfit changes exactly one of the three — that is a similar, and it is rejected. Getty also rejects near-duplicates as duplicate content in its general collection.
This is the rule worth internalising even if you never shoot for Getty, because it converts a subjective judgement into a checklist you can apply before you press the shutter, and it is stricter than what Adobe and Shutterstock describe in words.
The keyword side of the same problem
Metadata rejections
These are the most fixable rejections on the list, and at Shutterstock they are among the few you are formally allowed to correct and resubmit.
Titles
Shutterstock is the strictest and the most explicit. A commercial title must be in English, be a complete descriptive sentence of at least five words, accurately describe everything in the frame, and exclude special characters. It specifically warns against single words separated by commas, "because this looks like a list" rather than a title. If your workflow pastes your keyword string into the title field, every file you submit is failing this rule.
Adobe rejects files for inappropriate or irrelevant titles and keywords, and its title field will not take commas at all.
Keywords
Shutterstock names two keyword failures: repeated words or phrases — with the reassurance that "search still works with single entries" — and keywords that are irrelevant or incomplete. Adobe treats irrelevant keywords as a refusal reason and, for AI content, refuses anything that names a copyrighted artist in the prompt, title or keywords.
There is one cross-agency keyword failure that no single agency documents, because it only exists if you work across several: shipping Getty's controlled-vocabulary syntax to an agency that has no controlled vocabulary. Train - Vehicle and One Man Only are correct Getty terms and meaningless to an Adobe or Shutterstock buyer, who searches for train and man. Those terms are not usually rejected outright — they just quietly fail to match anything.
Metadata rejections are the ones worth automating away: an English sentence of the right length, keywords that are relevant and non-repeating, and the right vocabulary per agency. PixTagger writes all three from the image itself. First 15 files free.
Adobe Stock keyword generatorQuality and technical rejections
Adobe publishes the most detailed technical guidance, and the causes it names are worth reading as a pre-flight checklist rather than a post-mortem.
- Noise. The most common problem Adobe names. It points out that many cameras are excessively noisy at ISO 1600, and some at lower settings — which is a lower ceiling than most photographers work to.
- Artifacts from post-processing. Invasive editing produces them as reliably as a bad sensor does. Heavy noise reduction, aggressive sharpening and strong local adjustments all qualify.
- Sensor dust. Still a named rejection cause in 2026, and still the easiest one to prevent.
- Rolling-shutter artifacts in video. Adobe names skew, jello and flash banding specifically, and says clips containing flash banding or wobble will not be accepted.
Shutterstock's technical rejection list runs wider and covers several things that are not about optics at all: white balance, file integrity (corrupt files and transmission errors), resolution and aspect ratio — with pillarboxing and letterboxing accepted only for vintage content — image or video orientation, clip length outside 5 to 60 seconds unless the clip loops, unacceptable audio, and VR, 360 and 3D formats, which it is not currently accepting at all.
Legal rejections, which you cannot argue with
These are absolute. No amount of quality gets a file past them, and they are the reason a technically perfect image can be refused.
| Problem | What it covers |
|---|---|
| Intellectual property | Logos, trademarks, brand names, identifiable product packaging, and commercial products with distinctive designs as the main subject |
| Third-party copyright | Art, sculpture, street art, illustrations, fonts and graphic elements created by someone else — including in the background |
| Releases | Recognisable people, private property and artwork need valid releases. Adobe names incorrect or inconsistent information, illegible or incomplete signatures, and the same release reused across different models |
| AI labelling | Content mislabelled as to whether generative AI was used, or generated using a copyrighted artist's name in the prompt, title or keywords |
| Personal branding | Shutterstock rejects names, signatures, watermarks and datestamps burned into the file |
| Duplicate ownership | Content already submitted under a different contributor account — Shutterstock treats submitting material you do not own or control as grounds for account termination, not a procedural slip |
Note how the AI row differs by agency, because the policies are not converging. Adobe accepts AI-generated content if it is correctly labelled. Shutterstock and Getty do not accept contributor AI submissions at all. And Alamy went further on 1 September 2026 — AI content is now a material breach of its contributor contract rather than merely disallowed.
What to do with a rejection — and the rule that catches people
The single most important thing Shutterstock publishes about rejections is not a reason at all, it is a limit: "a second rejection is final, and the asset will no longer be eligible for submission." You get one corrected resubmission. Firing a file back untouched to see if a different reviewer disagrees is how you permanently lose it.
Shutterstock's correction route covers a specific list: metadata problems flagged as Commercial or Editorial Requirements, Translation Required, missing or incorrect model and property releases, unnecessary releases, and an invalid case number. Release problems specifically eligible for correction include a missing model name, a corrupted form, an adult form used for a minor, a missing signature and missing contact information.
Everything else is closed. Content previously rejected will generally be rejected again, approved content cannot be overwritten (delete it first), suspended content must not be resubmitted, and an identical file already pending review will be refused as a duplicate submission.
- Read which bucket the reason falls in. Metadata and release problems are correctable. Quality, similarity and legal problems are not — the file is done.
- Fix the cause, not the file. A Similar Content rejection on frame three of five is telling you something about frames one, two, four and five as well.
- Never resubmit unchanged. One correction is all you get.
- Check your own portfolio before uploading. Shutterstock's own advice for avoiding similar-content rejections is to look at what you have already had approved. Most similars are rejected against your own back catalogue, not somebody else's.
In short
Sources & further reading
- Adobe Stock — Reasons content is rejected
- Adobe Stock — Quality and technical standards
- Adobe Stock — Similar content versus spam
- Adobe Stock — The review process
- Shutterstock — Content review (full rejection reason list)
- Shutterstock — Why was my content rejected for Similar Content?
- Shutterstock — Why was my content rejected for title or keywords?
- Shutterstock — Content resubmission guidelines
- Shutterstock — Submitted to a different account or previously submitted content
- Alamy — Contributor contract changes (AI content as material breach)
Frequently asked questions
- Why does my content keep getting rejected for Similar Content?
- Because the agencies measure similarity more tightly than photographers do. Adobe asks for no more than three significant versions of one scenario and names sequential shots from the same moment as the failure case. Shutterstock counts the same subject with only a framing, crop or aspect-ratio change — and, for video, the same subject shot at a different clip length or with a different camera movement. Getty requires two of Models, Location and Wardrobe to differ.
- Can I resubmit a rejected stock photo?
- Sometimes, once. Shutterstock allows a correction for metadata problems (Commercial or Editorial Requirements), Translation Required, missing or incorrect releases, unnecessary releases, and invalid case numbers. Quality, similarity and legal rejections are final. And its guidelines state plainly that a second rejection is final and the asset becomes permanently ineligible — so never resubmit a file unchanged.
- What causes an Image Quality rejection?
- Adobe names noise as the most common cause, and points out that many cameras are excessively noisy at ISO 1600 and some below it. The other named causes are artifacts from invasive post-processing, sensor dust, and for video the rolling-shutter family — skew, jello and flash banding, which it says will not be accepted.
- Why was my technically perfect photo rejected?
- Most likely a legal reason, which quality cannot overcome: a visible logo, trademark or distinctive product design; third-party artwork, sculpture, street art or fonts in frame; a missing or defective model or property release; or an AI labelling problem. Shutterstock also rejects files with names, signatures, watermarks or datestamps burned in.
Written by a working stock contributor
NoSystem Images
Getty Images / iStock exclusive contributor since 2007
PixTagger is built by NoSystem Images, an exclusive Getty Images and iStock contributor since 2007, with a live portfolio of over 57,000 photos and 9,700 videos. Every keywording rule in the app comes from nearly two decades of actually selling on Getty, iStock and Adobe Stock — not from guesswork.
Related guides & tools
- Keywording similar shots without cannibalising yourself — the metadata half of the similar-content problem
- 7 Adobe Stock metadata mistakes — the metadata failures that never bounce, they just never rank
- Adobe Stock keyword generator — relevant, non-repeating keywords and a title that passes review
Stop hand-keywording every upload
PixTagger writes buyer-focused titles, descriptions and marketplace-ready keywords for your photos and videos in seconds — with a Getty controlled-vocabulary CSV, an Adobe CSV, and qHero export built in.