Storyblocks sets a keyword floor of six and a ceiling of 100, but the count that actually decides whether a clip gets found sits inside that range: the contributor help centre recommends 15 to 30, and says as many as 50 can contribute to a clip's success. The rules that send files back for editing are narrower — English only, no trademarks, and a title that is genuinely unique.
The keyword counts Storyblocks publishes
Three numbers appear in the Storyblocks contributor documentation, and they do different jobs. Six is the hard minimum — a file with fewer keywords does not meet the metadata submission requirements. One hundred is the hard maximum. Between them sits the recommendation: 15 to 30 keywords per file, which is the range the help centre points contributors to.
The keywording guidance adds a second framing of the same band. It tells contributors to aim for a minimum of 15 and says up to as many as 50 can contribute to a clip's success. Read together, that is a working range rather than a target. Fifteen is roughly where a clip stops being under-described; past about 50 you are adding terms unlikely to be the ones a buyer types.
- Minimum: 6 keywords per file.
- Recommended: 15 to 30.
- Still useful, per the keywording guidance: up to 50.
- Maximum: 100 keywords per file.
Nothing in the documentation rewards reaching 100. Storyblocks states that padding a list with unrelated words keeps a clip from reaching the right buyers and will not help your earnings. The cost of over-keywording is dilution rather than a rejection notice, which is why it goes unnoticed for months.
One concept per keyword, with a documented exception
The structural rule on Storyblocks is that each keyword should work as a single-concept search term. Most agencies share that instinct, but Storyblocks writes the exception down as well: multi-word terms are fine when the words are genuinely connected, and the documentation offers New York and computer screen as examples.
The distinction matters most when you convert a title into keywords. A woman walking a dog through an autumn park is not one keyword. It becomes woman, walking, dog, autumn, park — plus everything the title does not carry. Storyblocks tells contributors to start with the words from the title as individual keywords, then add related nouns, verbs and adjectives that describe the clip.
The documentation asks for four layers: subject, setting, action, and connected concepts. That last layer is the one contributors skip. The example Storyblocks uses is a birthday clip that should also carry fun — the abstract term a buyer searching for a feeling will type, which no literal description of the frame would produce.
Order a clip's keywords by how much work they do
Storyblocks accepts 6 to 100 keywords per file and recommends 15 to 30, covering subject, setting, action and connected concept. Reorder this sample clip's terms to the ten you would keep if the list had to be cut.
Titles are where most metadata resubmissions start
Storyblocks treats the title as a descriptive sentence rather than a keyword list, and the requirements around it are more rigid than the keyword rules. Three of them account for most editing requests.
Uniqueness comes first. Every file needs its own title, and the documentation is specific about what that means: you must change at least one word so the complete title is not identical to another file's. That applies across asset types and explicitly covers near-identical clips, such as animations with minor changes. Storyblocks suggests differentiating on clip length, colour scheme, shapes, or angles and actions.
Length is a range with a target inside it. The minimum is 16 characters, the maximum is 140, and the documentation says to aim for about 65 including spaces. Sixty-five characters is roughly one full descriptive clause — enough for subject, action and setting without padding.
The third rule is the one that catches bulk uploaders: the title must differ from the filename, and it cannot match the filename even with the extension stripped. Any workflow that copies a descriptive filename straight into the title field will come back for editing.
Storyblocks also asks contributors to spell-check both titles and keywords before submitting, and says it may ask you to edit and resubmit metadata that does not meet the submission requirements. That is a resubmission loop rather than an outright rejection, but it costs the same time and it holds the file out of the library while it runs.
English only, and no trademarks anywhere in metadata
The language rule carries no exceptions in the documentation: all titles and keywords must be in English. There is no secondary-language field to populate, which sets Storyblocks apart from agencies that accept or machine-translate local-language metadata. If your pipeline keywords in another language first, the translation has to happen before delivery rather than after.
Trademarks are barred from both titles and keywords, and the rule reaches further than logos. Storyblocks includes brand names that are commonly used generically — the ones contributors type without thinking of them as brands at all. A clip of someone drinking from a paper cup does not get the coffee chain's name, and a clip of a vacuum cleaner does not get the brand that has become a verb.
Three ways metadata gets in, and which one scales
Storyblocks accepts metadata through three paths, and the one you pick decides how much of this work stays manual.
The web form is available for every asset type. Contributors edit titles and keywords on the online platform under the content tab. It is the only option for audio artists, and it is workable for a handful of files at a time.
CSV upload covers footage, animation and templates; audio artists cannot attach a CSV yet. The detail most contributors miss is that Storyblocks accepts CSV files formatted for Shutterstock or Pond5 without any editing. If you already build a Shutterstock CSV, the same file can go to Storyblocks as it stands.
Embedded IPTC Core data works for video only. Storyblocks reads the Headline, Description, Title and Keywords fields from the file itself, and the order it checks them matters: to populate the site's Title field it looks at Headline first, then Description, then Title. Writing a clean Headline is the most reliable way to control what ends up as the published title.
Delivery scales along the same line. The web uploader takes up to 100 files per session, and larger batches go over sFTP. Release forms are a separate step and go through the Contributor Portal rather than FTP.
Build a Storyblocks metadata row
Storyblocks accepts a metadata CSV for footage, animation and templates, and takes Shutterstock- or Pond5-formatted files without editing. Limits shown are its own: titles 16 to 140 characters, 6 to 100 keywords, English only.
A keywording pass that holds up on review
Put the published rules in the order you would actually work through them and the pass is six steps.
- Write the title first, as a sentence, and land it near 65 characters. Confirm it is not the filename and not a title you have already used.
- Split the title into single-concept keywords, keeping genuinely connected multi-word terms intact.
- Add the layers the title does not carry: setting, action, mood, time of day, season, and the abstract concept a buyer might search instead of the literal subject.
- Count. Under 15 and the clip is under-described. Over 50 and you should be cutting the terms you cannot defend as something a buyer would type.
- Run a trademark pass over both fields, including brand names that have slipped into generic use.
- Spell-check both fields, then confirm every term is English.
Storyblocks also points contributors to its in-demand dashboard and says to use any relevant terms from it. That is the one point in this process where the agency tells you what buyers are searching for rather than asking you to guess, so it belongs in the pass rather than in a separate research session.
Where the documentation stays silent
Two gaps are worth naming, because contributors fill them with assumptions carried over from other agencies. The first is keyword order. Storyblocks publishes counts and quality rules but gives no guidance on sequence, so there is no documented reason to believe the first five keywords carry more weight than the last five. If you are front-loading your strongest terms, you are applying another agency's habit here, not a Storyblocks rule.
The second gap is the description field. The submission requirements page sets out title rules in detail and keyword rules in detail, but does not publish separate length or style rules for descriptions, nor a category taxonomy. For video, the embedded Description field is read and can fall back into the site's Title, which is the only documented behaviour tying the two together. Treat the description as supporting prose rather than a second keyword field.
Neither gap is a licence to improvise. The published rules are specific enough that most resubmissions trace back to one of them being skipped rather than to anything undocumented: a duplicate title, a title lifted from a filename, a non-English term, a trademark, or a list that never reached fifteen keywords.
Running Storyblocks alongside everything else
Storyblocks overlaps heavily with the other footage marketplaces but differs in exactly the details that break automation: the 140-character title ceiling, the filename check, English-only metadata. The full field reference covers every submission field, and the guide to Storyblocks rejection reasons covers what happens when a file comes back instead of going live.
Holding each agency's version of these limits in your head across a nightly batch is where the time actually goes. Rastock AI writes titles, descriptions and keywords against each agency's own rule set and delivers over FTP or SFTP to 10+ agencies, with banned and mandatory keyword handling per destination and rollback if a batch goes out wrong. You can see what it covers and how it is priced. There is a 14-day free trial and no card is required to start.
Frequently asked questions
How many keywords does Storyblocks require?
Storyblocks requires a minimum of six keywords per file and allows a maximum of 100. Its contributor documentation recommends 15 to 30 and notes that as many as 50 can contribute to a clip's success. Six will pass the submission check, but a clip with six keywords is unlikely to match many buyer searches.
Can Storyblocks keywords be more than one word?
Yes, when the words form a single connected concept. Storyblocks asks contributors to treat each keyword as a single-concept search term and gives New York and computer screen as acceptable multi-word examples. A full phrase describing the whole clip is not a keyword — split it into the individual nouns, verbs and adjectives instead.
Why does Storyblocks ask me to change my title?
Most often because the title is not unique or it matches the filename. Storyblocks requires at least one word to differ so no two complete titles are identical, including across near-identical clips, and the title cannot match the filename even with the extension removed. Titles also have to fall between 16 and 140 characters.
Does Storyblocks accept keywords in languages other than English?
No. Storyblocks states that all titles and keywords must be in English, and there is no separate field for other languages. If your workflow generates metadata in another language, translate before delivery. That differs from agencies which accept or machine-translate local-language metadata, so a shared multi-agency CSV needs an English column for Storyblocks.
Can I reuse my Shutterstock CSV for Storyblocks?
Yes, for footage, animation and templates. Storyblocks accepts CSV files formatted for Shutterstock or Pond5 without any editing. Audio contributors cannot attach a CSV. The metadata itself still has to satisfy Storyblocks' own limits: six to 100 keywords, titles of 16 to 140 characters, English only and no trademarks.
Are trademarked words allowed in Storyblocks keywords?
No. Storyblocks bars trademarked names from both titles and keywords, and explicitly includes brand names that are commonly used generically. That covers terms contributors often type without thinking of them as brands. Removing them is a separate pass from spell-checking, because a spell-checker will not flag a correctly spelled brand name.