vs. Kanban

Kanban is probably the most honest of the traditional approaches about what it actually optimizes: the flow of delivery. And that's exactly where its blind spot is.

The missing metric

Kanban tracks Lead Time (from request to delivery) and Cycle Time (from start to delivery). These are useful numbers. They tell you how fast work moves through your pipeline.

But they never tell you if the work mattered.

You could have a perfect Kanban board. Low cycle time, smooth throughput, minimal WIP. And still have no idea whether you're learning anything. The board ends at "Done." In LDD, shipped isn't done. Validated is done.

LDD introduces a metric Kanban never had: Learning Time. From hypothesis to validated decision. That's the number that actually tells you if you're making progress or just moving cards.

WIP limits vs. exposure limits

There's an interesting structural parallel. Kanban limits work in progress to optimize delivery flow. LDD limits exposure in progress to optimize learning clarity. Same idea, different thing being optimized.

In Kanban, too much WIP slows delivery. In LDD, too much unvalidated exposure makes it impossible to read the signal. One slice, one belief, one answer.

Beyond delivery

Kanban is a delivery system. It's very good at what it does: making work flow smoothly from "requested" to "done." But it has no opinion on what happens after delivery. Did the user care? Was the hypothesis right? Should you double down or abandon?

LDD picks up where Kanban stops. The loop doesn't end at delivery. It ends at a decision: double down, pivot, or abandon. Delivery is just the midpoint.

Can they coexist?

Yes. You could use a Kanban board to manage the Build phase of a learning loop. The flow optimization still has value when you're moving work through execution. But the board needs to extend past "Done" into "Validated" and "Decided," or you're only seeing half the picture.