Once your library grows to hundreds of creatives, finding versions, replacing aging assets, and granting collaborator access all become a burden. This guide explains naming and grouping, stage-based archiving, fatigue signals, and collaboration permissions.
As the number of ad accounts grows, creative management problems often appear before media-buying strategy problems do. The issue is not that the creatives are bad, but that they become hard to find, it is unclear which ones are running, and replacing them at the right time becomes difficult. Once the library reaches hundreds of assets, these small frictions steadily consume time every day.
Lock down naming and grouping first
Naming chaos is the source of most problems that follow. Pick a fixed format you will not forget, such as market + product line + creative type + version + date, for example US_Lighting_ShortVideo_v3_20260914. Organize folders in three levels by market, product line, and creative type. Do not make the hierarchy deeper than is practical, or people will stop clicking through to find things.
Group by purpose, not by source. Grouping by source—who designed it or where the inspiration came from—may look tidy, but it is not useful during daily campaign work. Grouping by purpose—which asset is being tested, which is in a primary slot, and which has been retired—is what you actually need to review every day.
One point deserves special emphasis: do not put account names in file names. Reusing the same creative across multiple accounts is normal. Adding account names causes the same file to be stored several times, and once versions multiply, they start overwriting or conflicting with one another.
Archive by stage: testing, primary, retired
Creative assets do not need a complicated lifecycle. Three stages are enough.
Testing assets are the ones launched within the last week or two. Keep this area small so assets can move in and out quickly, and define the evaluation criteria in advance; otherwise they may remain in testing indefinitely without a conclusion. Primary assets are the few creatives delivering stable volume. Any change must create a new version: back up first, then edit, and keep a record of who changed what. Retired assets are those that have shown sustained declines or have become restricted; keep them archived but stop running them.
Do not rush to delete retired assets. When seasons, markets, or audiences change, modifying an older creative and running it again is often faster than building a new one from scratch. The reusable value is usually in elements such as the structure and spoken-message rhythm.
What creative fatigue looks like and when to replace assets
Fatigue rarely happens all at once. Usually several visible signals appear together: frequency rises while click-through rate and conversions both fall; cost per click climbs for several consecutive days even though the audience and bid have not changed; impression share within the same audience is already high but conversions no longer rise with it; or comments start saying people are tired of seeing the ad.
Act when two or more signals appear; do not treat a single fluctuation as fatigue. Keep two or three creatives ready to rotate in primary placements, while continuously introducing new assets in testing slots, so you have replacements when a primary creative loses volume. Start with small changes—switch the first frame, change the opening copy, or move the call to action—instead of rebuilding the entire creative at once, otherwise you will not know which change made the difference. Compare data from the same time window; comparisons across different periods are not meaningful.
Versions and permissions in multi-person collaboration
Collaboration problems usually cluster in three places: someone changes copy without syncing it, two people edit the same asset at the same time, or a person leaves but their access remains active.
A relatively simple approach is to separate read and write permissions. Give media buyers read-only access to the creative library, keep write access with the creative owner, increase version numbers instead of overwriting files, and record who used which version and when before launch. Maintain the mapping between creatives and ad accounts in a separate table rather than leaving it only in one person's memory.
There is also an account-safety layer: when multiple ad accounts use exactly the same creatives, launch at similar times, and target the same audience, those factors together can form behavioral linkage signals. That is why creative differentiation—at least one difference in visual style or copy structure—and staggered launch times should be used together. Operating multiple accounts in their own fixed environments is the prerequisite for making this rule workable. Tools such as PurpleMark provide separate environments and collaborator permissions by account.
How to handle rejected creatives
After a rejection, the worst approach is to change two words and resubmit essentially the same asset, because it may be treated as a duplicate submission. First identify the category of the rejection: whether the creative itself contains noncompliant wording or inconsistent promises, whether the landing page does not match the creative, or whether the product category itself is restricted. Then fix the specific issue, and make the change materially different.
Move rejected assets into a separate folder and label the rejection reason and date so an uninformed teammate does not pick them up later and submit them again. If the same account receives repeated rejections, pay attention to the effect on the account's overall performance. In that situation, changing the way the message is expressed is often more efficient than repeatedly forcing revisions to the same creative.
The final step before launch
Put creative review status, account-environment checks, and mapping-table updates on the same checklist, and complete them all before clicking publish. That takes far less effort than later tracing who used a creative and which environment it ran in.
