Schema Ownership Determines Integration Power
Sara Du, co-founder and CEO of Alloy, on iPaas vs. universal APIs
Whoever controls the schema controls what developers can and cannot build. A universal API does not just pass data through, it decides that an order, a fulfillment, or a delivery should be represented in one specific way, and that choice can hide important differences between platforms. That is why vertical players often start stronger, because they know which fields matter in the daily workflow of one industry and can keep updating that model as the industry changes.
-
In commerce, the hard part is domain mapping. Shopify, Amazon, WooCommerce, NetSuite, and other systems all store products, orders, customers, and fulfillments differently. A unified schema has to flatten those differences into one model, which makes it faster to integrate, but always leaves edge cases outside the common abstraction.
-
That tradeoff is why universal APIs often cover the first layer of read and write use cases, then customers move up to embedded iPaaS when they need custom fields, user specific logic, approvals, or configurable workflows. The schema gets a product live quickly, but real production integrations usually need knobs the standard model does not expose.
-
Vertical focus improves schema quality because the company stays close to one ecosystem and its partner conventions. Horizontal focus improves distribution because the same commerce data can power fintech underwriting, ERP sync, and other cross industry use cases. Alloy started in commerce, then broadened its positioning for exactly that reason.
The next wave of universal API companies will compete less on building connectors and more on owning the best data model for a workflow, then layering configuration, native escapes, and higher level products on top. As AI makes connector generation cheaper, the durable advantage shifts to schema judgment, partner access, and knowing when one vertical model can stretch into adjacent markets.