Operations

Promise gap

The difference between the delivery time a customer is promised at checkout and the time the operation actually takes. A wide gap costs in both directions: promise too long and the order is lost at checkout, promise too short and it breaks at the door. Closing it means calculating the promise from live capacity and real delivery history rather than a fixed buffer. Nash calibrates promises against what the network actually delivered, zone by zone.

Where this fits

The day-to-day vocabulary of running deliveries. Routes, waves, drops, manifests, ETAs, SLAs. These are the units operators reason in, and the inputs every downstream model needs.

Related terms

Promise buffer: The padding added to a quoted delivery time so the commitment survives a bad day. Every operation has one, most do not know its size, and it is paid for twice: once in customers who choose a faster competitor, and once in capacity held against a risk that rarely lands. Delivering early half the time is the sign the buffer is too wide.
Route adherence: How closely the executed route matched the planned one: sequence kept, stops served in window, no improvised detours. Low adherence usually means the plan was wrong rather than the driver was, because drivers route around what the model could not see. Measuring it is how an operation finds the constraints missing from its own planning.
Cost to serve: The full cost of fulfilling an order for a given customer, channel or zone: picking, packing, the delivery itself, failed attempts and redeliveries, returns, support contacts, and the margin given up to a marketplace or provider. Cost per delivery is one line of it. Two orders with the same delivery fee can cost different amounts to serve, which is why the decision about who carries each order belongs next to the promise made for it.
Hours of service: The regulated limit on how long a commercial driver may drive and work before a required break or reset. A hard routing constraint: a plan that ignores it produces routes a driver cannot legally finish, and the break has to be placed where there is somewhere to stop.
Order density: The number of orders in a given area and time window. Higher density means more stops per route and less driving between them, which is the main reason the largest retailers deliver at a lower cost per delivery than everyone else. Density is partly given, by where customers live, and partly made: batching orders, offering slots that cluster demand, and pooling volume across stores or carriers all raise it.