Every season launch is a fire drill. Which photo was approved, which variant still lacks an on-model shot, who gave the go-ahead and who is silent. The answers live scattered across twelve Slack channels, six WhatsApp threads, and Dropbox folders full of _FINAL_v3_USE_THIS.psd.
Each one is real. Each one is solved. Not with another tool, but with a platform that takes the job seriously.
Twelve channels, three of them archived, one with the word "release-final" in its name. Nobody knows where the current state is actually being discussed.
Releases as their own object with linked assets, approvals, and logs. Not a channel name, but a database row with an audit trail.
"Looks good 👍" is not an approval. No date, no audit trail, no verifiable state. When the insurance asks 14 months from now, there is nothing to show.
Configurable approval flows with stages, any or all rules, defined approvers, and an append-only audit log. Append-only because compliance does not ask whether you have time right now.
shirt_v1.jpg. shirt_v2_FINAL.jpg. shirt_FINAL_FINAL.jpg. shirt_v3_USE_THIS_FINAL.jpg. Which file is live in the shop today? Nobody knows.
One library per account with SHA-256 dedup. Format converter, batch resize, background removal create new asset rows with tool chain provenance instead of overwriting the source.
Six assets, three status queries, one empty slot, one WhatsApp conversation from last week. With four drops per season, the status table becomes the bottleneck.
Approval state per asset, visible everywhere the asset appears. Kanban view across the entire account, filters per stage and per approver.
Three red cells in an Excel spreadsheet. Three variants without an on-model shot. You only see it when someone opens the sheet, which rarely happens.
product_assets links assets at the variant level via variantId. Releases show the readiness rollup per child product. What is missing is a list, not a gut feeling.
Without clear approval stages, a single silent approver delays the entire season. Nobody knows who they are waiting for, or since when.
Approval tasks bound to approvers, wait time per task visible. With rule all, every named approver blocks; with rule any, one is enough. Silence becomes a data point.
Akeneo, Pimcore, and Salsify target 50,000-SKU enterprises with twelve-month onboarding. KMUPIM targets DACH SMB brands with 10 to 200 SKUs that launch drop by drop, have no steering committee, and need a platform that works on day one.