We sell manufacturing execution to SME fashion brands in Europe and North America. Not a tool they operate. The work itself, run by agents.
The unit is a production run. A brand submits a design. The brand-side agent builds the tech pack, sets cost and timing targets, and holds the critical path. The supplier-side agent handles capability matching, quoting, sampling, QC and logistics. The two agents transact with each other, so the human handovers that stopped on-demand production from ever scaling are removed from the process rather than staffed around.
Revenue comes from the brand side and scales with brand maturity. Brands under £1M ARR enter on commission, up to 15% per order, no subscription and no commitment, worth £5K+ per brand per year. Brands at £1–10M ARR move to a usage-based subscription with a flat fee per run, worth £20K+ per brand per year. Brands above £10M ARR buy agents as a service, usage-based on top of a lower per-run fee, worth £100K+ per brand per year. Brands move up the ladder as they grow, so account value expands without new acquisition spend.
Suppliers join free. That keeps capacity liquid and gives us execution data on every order. We monetise them separately with agents that replace their account management function, worth £20K+ per supplier per year and a 20–40% uplift on topline.
The model compounds because every run is captured as structured production memory that trains the agents. Cost to serve falls while price holds. Gross margin is 65% today and expands toward 80%+ as automation replaces coordination. Retention comes from repeat production cycles, not one-off sourcing, and switching cost rises with every order because leaving means leaving your production history behind.
On-demand manufacturing never scaled because every step is a human handover. Makea puts an AI agent on the brand side and one on the supplier side, carrying a design to finished product. Patent-ready agentic architecture for textiles, trained on China's largest manufacturing dataset.