← Back to Stories

Story · Case study

One network review a week instead of one a quarter

Route economics used to be a quarterly project with a start date. It is now an operation the planning team submits on Monday and reads before lunch.

The company

Company
Kestrel Logistics
Industry
Logistics
Company size
1,001–5,000 people

The starting point

Kestrel Logistics runs 1,900 vehicles out of fourteen depots. The economics of that network — which routes earn their cost, which depot should take which drop, what a closure would do to the rest — were reviewed once a quarter by three analysts over about five weeks.

The review was thorough and the network moved faster than it. By the time a recommendation was agreed, two of its assumptions had changed: a customer had renegotiated, or a depot had taken on volume it was not sized for.

What changed

What they described

Kestrel's head of network planning described the review rather than the tool: what makes a route economic, how a drop is costed when two depots could serve it, what counts as a fair allocation of fixed depot cost, and which constraints are real (driver hours, vehicle type) versus habitual (a customer who has always been served from one depot).

Separating those two was most of the value, and it came out of writing the description down rather than out of anything the platform did. Several of the constraints the analysts had modelled for years turned out to be habits nobody had questioned since a depot opened.

The results

The scenario that paid for it

A depot lease came up for renewal at a rent Kestrel considered indefensible. The question — what happens to the network if we do not renew — had previously been a four-week study, which meant it was never asked until the lease was nearly expired.

It was submitted as model-depot-change on a Tuesday afternoon. The answer said the closure was survivable, that two neighbouring depots absorbed most of it, and that eleven routes became uneconomic in a way that would need renegotiating rather than absorbing. The negotiation that followed used the eleven routes as its opening position.

Example

Routes scored each week

1240

Cost per drop

-4.8%

against the same quarter last year

Routes flagged late-risk

31

Days to a network review

1 day

from five weeks

“A quarterly answer to a weekly question is a wrong answer three times out of four. That was the whole business case.”

— Declan Moss, Head of network planning, Kestrel Logistics (an example business)

The apps

What it cost to run and what to expect

What the planning team says about it

That the weekly rhythm matters more than the model. A slightly worse answer every Monday beats a better answer every quarter, because the network's questions arrive weekly and an answer that arrives after the decision is not an answer.

Connected to this story

Tell me when it's ready

We keep your email only to tell you once, when the app is ready, and delete it then.

What happens next

Thank you. We will write once, when it's ready.