Rebuilding Tractor At Home, the direct-to-home channel.

Tractor At Home ran on a chain of third-party tools — one for the storefront, another for scheduling, another for fulfilment notifications. Each worked, but none of them knew about the others.

Customers hit the gaps: order confirmations that didn't match delivery windows, changes that couldn't be made without calling, and support requests that bounced between the store team and a vendor's help desk. Every improvement we wanted was gated by someone else's roadmap.

I mapped the full customer journey — discovery, menu selection, checkout, delivery window, and post-order support — and identified where the handoffs between tools were causing failures rather than just friction.

We rebuilt the parts that defined the experience and integrated them with internal operations: menu and availability driven by the same source as the stores, order data flowing into the same reporting the operations team already used, and support visibility for the people actually answering customers.

Where a vendor did something well, we kept it — but behind our own interface, so a vendor change never became a customer-facing change.

Checkout got shorter and clearer, order and delivery information finally agreed with each other, and fewer support conversations required a handoff to a third party.

The bigger win was tempo: improvements shipped on our schedule instead of waiting in a vendor's queue, so small fixes stopped accumulating into big problems.

We improved the experience based on support tickets and observation before we had clean funnel analytics. The changes were right, but proving their impact took longer than it should have. Instrumentation should have shipped with the first release.