BaaS as Operational Infrastructure
Fintech Fastlane: The Unit Economics of the Banking-as-a-Service Toll Road
For Instacart, BaaS is a logistics tool, not a profit center. The card exists so a shopper can pay for the exact basket at checkout, with transaction level controls, instant issuance, and fraud limits that make the in store run faster and cleaner. That creates better fulfillment and retention, but it does not create the same built in push to maximize swipe volume that drives fintechs like Cash App or Klarna.
-
The practical workflow is simple. Instacart issues a shopper a card tied to a specific order amount, then approves or declines each purchase based on merchant, amount, and timing rules. Those controls became core product features for Instacart, Uber, and DoorDash because they reduce checkout friction and mistakes in the field.
-
That is different from a fintech card. Fintechs use issuing to move revenue off the front end and onto interchange, so more card spend directly means more gross profit. Embedded finance companies use the same rails mainly to make the main app work better, so transaction growth follows delivery growth, not a deliberate monetization loop.
-
This customer mix changes BaaS economics. Providers serving fintechs can land a few giant accounts with very high volume, but they take more concentration risk and more pricing pressure as those customers scale. Providers serving embedded finance get more distributed revenue and lower insourcing risk, but usually slower volume growth per customer.
The next wave of BaaS growth comes from more non fintech apps turning payments and cards into invisible infrastructure inside their product. That favors providers that can make issuing programmable, reliable, and easy to plug into operational workflows, because the biggest opportunity is helping software companies remove friction from real world actions, not just helping fintechs manufacture more card volume.