Modular content governance is the set of owners, approval scopes, usage rules, version and expiry dates and metadata that decide when an approved component or module can be reused, and where. Its central rule is that approving a module approves the module, not every asset it later lands in: each finished piece is still reviewed in its own context and, for a prescription drug or biologic, submitted to FDA on Form FDA 2253 when it is first disseminated or published.
Most modular programs do not fail at launch. They fail a year later, when the library holds modules nobody owns and rules nobody enforces, and writers quietly go back to copying from last year's detail aid. This article covers who owns each layer, what an approval at each layer covers, the rules a module should carry, versioning and expiry, the label-change cascade, and the metadata and metrics that keep a library trusted. It builds on our guide to building a claims library, reflects FDA regulations as of November 2026, and is not legal advice. Your regulatory and legal teams set the rules for your products.
Who owns components, modules and assets?
Ownership has to be set per layer, because each layer fails in a different way. Components are the smallest approved blocks: a claim with its reference, a statistic, a line of copy, an image, the ISI. Modules are approved combinations of components with the context they need. Assets are the finished pieces that reach an audience: an email, a banner, a detail aid, a video.
Components belong to the function that can vouch for their content. Medical owns claims and statistics, and the accuracy of the reference behind each one. Regulatory owns the ISI, the indication statement and the required qualifiers. Brand owns copy lines, imagery and the visual system. Legal owns what carries legal exposure: trademarks, comparative claims, and the licence and talent terms attached to images and footage.
A module needs one named owner, usually on the brand team, who answers for the combination and decides when it is retired. Content operations owns the library itself: the taxonomy, metadata quality, expiry jobs and reports. The asset owner is whoever disseminates the piece, and that person owns its first-use date, its 2253 submission and its withdrawal.
Keep owners and approvers separate in the record. Medical may approve a module that Brand owns. Confusing the two is how modules end up with nobody responsible for retiring them.
What does approval cover at each layer?
Approving a component confirms that the wording is accurate, supported by the cited reference and consistent with the label. Approving a module confirms that the components work together, with the required context, for the audiences and channels named in its rules. Neither approval sees the finished piece.
Asset approval is where net impression is judged: the headline beside the claim, the image beside the statistic, the space and time given to risk, and how several approved modules weigh against each other on one page. Two modules that are each balanced can sit together in a layout that is not. That is why MLR still reviews every asset. The useful agreement with reviewers is about what they check, not whether they check: for reused modules in an approved template, review concentrates on the seams, meaning what is new, what sits next to what, and whether the rules were followed.
The FDA obligation follows the same logic. Under 21 CFR 314.81(b)(3)(i), an applicant submits specimens of promotional labeling at the time of initial dissemination and advertisements at the time of initial publication, each with a completed Form FDA 2253 and the current professional labeling. For biologics, 21 CFR 601.12(f)(4) applies the same requirement. What goes to FDA is the piece the audience sees.
So one module, approved once, can sit behind many separate 2253 submissions, one for each piece that uses it, and a revised piece is a new submission. FDA's electronic submissions guidance also describes a withdrawal request for a piece that was submitted and will never be used. Our article on Form FDA 2253 covers the filing detail.
What usage rules should each module carry?
A module without rules is a sentence with a logo. The rules let a writer use it without asking, and let a reviewer trust that it was used as approved. Every module should state its allowed audiences, channels and markets, its required pairings, its prohibited combinations and any placement rules.
An illustration. A fictional brand, Brevanta, has an efficacy module built from its primary endpoint claim, the study population qualifier and a chart. Its rules say: HCP audiences only. Approved for email, detail aid and congress panels, not paid social. Must appear with the HCP ISI at its current version and the full indication statement, with risk information on the same page or screen. Cannot appear beside the secondary endpoint module from a different trial, because the layout would imply a comparison the data do not support.
Rules should be structured fields wherever possible, not a note in an attached PDF. A rule a system can check, such as a required pairing or a banned combination, gets enforced at assembly. A rule written in prose gets enforced when a reviewer happens to remember it. Veeva Vault PromoMats supports this structure: its help documentation describes Rulesets, collections of business rules that give instructions on how to use a content module's assets and can be shared across modules.
How should versioning and expiry work?
Version every layer, and have each module point to specific component versions rather than to whatever is latest. When Medical updates a claim, the claim gets a new version and every module that uses the old one is flagged for re-approval. Silent updates are the fastest way to lose a reviewer's trust, because the module they approved is no longer the module in the library.
Give each module two kinds of date. A periodic review date, set in your SOP, forces someone to confirm the module is still accurate and still used. Hard expiry dates come from its components: an image licence term, talent rights on footage, a reference that a newer data cut will supersede, a label version. The module expires on the earliest of them, and the record should show which one drove it.
Expiry has to stop new use and trigger a decision about live use. Blocking new use is the easy half. The harder half is the assets already in market that contain it, which keep running in email platforms, on websites and on rep devices after the library record changes. Veeva describes automatic expiry and single-click withdrawal in PromoMats, including automatic withdrawal of obsolete digital content when it expires. That covers what the system publishes. Every other channel needs an owner and a confirmed removal date.
What happens to modules when the label changes?
Under 21 CFR 314.70(a)(4), an applicant must promptly revise all promotional labeling and advertising to be consistent with a labeling change, and 21 CFR 601.12(a)(4) says the same for biologics. In a modular program, that duty runs as a cascade from the bottom up.
Start with the label sections that changed. Find every component tagged to them: the claim drawn from the changed clinical studies table, the ISI, the dosing statement. From each component, walk up to every module that uses it, and from each module to every asset that uses the module, whether it is in market or in review. Veeva PromoMats offers a Where Used report that tracks where claims are used. Whatever the tool, the walk only works if the links were recorded when the asset was built.
Then fix in the same order. Update and re-approve the component once. Re-approve the modules that use it, or retire the ones the change breaks. Withdraw assets that now contradict the label, and send the rest back through review as finished pieces, each on a new 2253 when it goes back into use. Our article on label changes and live materials covers the withdraw-or-revise call and the field channels that are easiest to miss.
What metadata and metrics keep a module library healthy?
A module nobody can find gets rewritten, and the rewritten version skips the library. Tag every component and module from controlled lists, not free text: product, indication, market, audience, channel, claim type, label section and label version, reference IDs, owner, status, approval date, review date and expiry date. A search for an HCP efficacy module approved for email should return a short list, not a hundred near-duplicates.
Veeva's documentation shows the starting point. A content module record holds the product, country, custom fields and IDs, can be related to several products and countries, and can group its assets into combinations for specific channels, messages or personas. Fields such as audience, label version and expiry are worth agreeing before the library grows, because retrofitting tags onto an established library is a project nobody funds.
Then measure the library, not just the review queue. Reuse rate: the share of new assets built mainly from approved modules. Orphaned modules: approved but used in no asset over a set period, which usually means the module does not fit the briefs or nobody can find it. Expired in market: live assets that contain an expired module or component, which should be zero. Duplicates: near-identical modules for the same claim. Label-change closure time: days from a label change to the last affected asset fixed or withdrawn.
Read the numbers together. If reuse is low and orphans are high, the library is being bypassed, and governance has failed whatever the SOP says.
Where SwishX fits
Governance holds when every asset carries its own record of which claim, reference and safety information it used. Magic Studio generates video, visual aids, approved email, social ads and web content from one Brand Dossier and claims library. Every claim is cited to its source and checked against the label for its market, fair balance and safety information are placed in the asset as it is made, and anyone can ask an asset where a line came from, which is the question a label change asks of every piece in market.
SwishX is not a replacement for Veeva PromoMats or your MLR review. It sits upstream: finished work goes into the MLR queue and on into Veeva PromoMats with every claim and its reference attached, and your reviewers still approve each piece in context. See how it works in Magic Studio.
Frequently asked questions
What is modular content governance?
Modular content governance is the set of owners, approval scopes, usage rules, versions, expiry dates and metadata that control how approved components and modules are reused in finished assets. It decides whether a module library stays trusted by writers and MLR reviewers or gets bypassed.
Does approving a content module approve every asset that uses it?
No. A module approval covers the combination for its named audiences and channels. Each finished asset is still reviewed in context, because layout, imagery and neighbouring modules change the net impression, and for prescription drugs each disseminated piece is submitted to FDA on Form FDA 2253.
Are content modules submitted on Form FDA 2253?
Form FDA 2253 accompanies promotional labeling and advertising at initial dissemination or publication, under 21 CFR 314.81(b)(3)(i) and, for biologics, 21 CFR 601.12(f)(4). In practice the submission is the finished piece that uses a module, with the current prescribing information, and your regulatory team confirms the approach.
Who should own a content module?
One named owner, usually on the brand team, who answers for the combination and decides when it is retired. Medical, regulatory and legal approve their parts, and content operations owns the library's metadata, expiry jobs and reporting.
How often should content modules be reviewed?
FDA does not set a review interval for content modules. Companies set a periodic review date in their SOP, and expire a module early when the label, a supporting reference, or an image or talent licence changes.