Case Studies, Published Only When Verified

What every Agentis case study will measure

Each study will follow the same structure so results can be compared across merchants. The headline metric is net margin recovered over a defined period, usually a Q4 or a major promotional window, expressed as a percentage of revenue and reconciled against the merchant's own P&L rather than a dashboard estimate. Supporting metrics cover the share of orders that were flagged or held before fulfillment, the average margin gap on those orders, and how many held orders were released, adjusted, or cancelled, because a profit floor that costs good orders is not a win.

Every number will be reviewed by the customer's finance lead before publication, and the evidence behind it is the Evidence Ledger entry Agentis writes for each order decision: the live COGS, freight, fee, and FX inputs, the floor that applied, and the outcome. Where a merchant asks not to be named, the study will be published with the vertical and GMV band instead, and it will say so.

Until a customer result is verified, the closest public evidence is the ecommerce benchmark library and the margin leak cost ladder, which document the leak types these studies quantify.