Field Notes · A Project, Start to Finish

Go Live in Fifty Five

A vehicle importer bound to century-old paperwork, a CFO who wanted to see the return, and a promise that sounded like a fairy tale: live in fifty-five days. “What — can this really happen?” asked Jenny.

“What — can this really happen?” Jenny kept the invoices at one of the country's largest vehicle importers, and she had learned not to believe promises. The company brought in some 30,000 vehicles and more than 50,000 spare-part SKUs a year, on import routines inherited — almost literally — from the Ottoman and British administrations that drew the region's borders a century ago. This is the story of what happened when it decided to stop.

The problemThe paperwork empire

There was no real way to track a shipment's status; the systems were manual and out of date; everything leaned on the person who happened to know. Setting up a new shipment was a small “project” in itself. And the finances were worse: invoices arrived in euros, dollars and shekels, each to be checked by hand against a tariff someone had — or hadn't — negotiated for that deal, exchange rates reconciled, VAT extracted, customs costs keyed in one line at a time. Classify a gasket wrong — a safety part in composite zinc rather than graphite, eligible for a tax break, but made in China — and the error compounds downstream. Worst of all, the weight the storage charge was based on would turn out not to match the weight on the bill of lading. Errors ran to five figures, in dollars. “Will we forever eat bitterness?” one of the team asked.

A row of colourful toy cars, standing in for the vehicles at the heart of the import business
The toys of the trade. Thirty thousand vehicles a year — and behind each one, a shipment, a tariff, a stack of invoices in three currencies.

The testShow me, don’t tell me

The CFO set the bar: “We want to see the ROI — how a system really brings money home. What is it, an ATM?” Fair question. So the import manager built a real test case — real shipments, real invoices, real exchange rates, real tariff models — and put it to the vendors. One answered with slides: our system will audit x% of invoices; here is clear, indisputable ROI. The other refused the slides outright: we won't pass you papers built on someone else's data — we'll stand up a real system and run yours. They did: ten of the importer's own shipments, eight by air and two by sea, streamed into a live environment, the audit robot run against the actual rates, deviations surfaced at the level of each individual charge.

“What is essential is invisible to the eye.”

— Antoine de Saint-Exupéry, The Little Prince

You cannot judge a system by the canvas its salespeople paint. The important things are hidden — in the details, in the data, in whether a process actually changes. Only the empirical run told the truth.

High-bay warehouse racking filled with stock, representing a 50,000-SKU spare-parts operation
Fifty thousand SKUs. Behind the vehicles, the parts business — where a promise like “the gasket will be with you in four business days” depends on knowing exactly where every shipment is.

Success criteriaThe four corners of a project

Before a line of it was built, the project was held to four plain principles — the ones that decide whether any implementation succeeds:

  • Content — every agreed requirement met; short of it, there's no value for money, and no ROI.
  • Time — the schedule held, as a threshold condition; a slipping project bleeds return and ties up both teams.
  • Budget — kept; overrun hurts the customer, and scope-creep ties the supplier to a project waiting on cash flow, which erodes the whole solution.
  • Both win — customer and supplier satisfied at the end. If either walks away damaged, the project failed, whatever the status report says.
Diagram: project success criteria — budget, time frame, content/SOW arranged around a 'both win' centre
The four corners. Content, time and budget around a single centre that too many projects forget — that both sides come out ahead.

The critical phaseFit and gap

The make-or-break stage has an unglamorous name: Fit & Gap Analysis. Map the real processes as they are; define the change against new standard ones; identify every business rule the system's robots must know — until the whole organisation, shippers included, speaks one language. (“We all speak SCMaster-ese now,” the import manager would say.) Its output, the Scope of Work, becomes the reference model every later phase is measured against — the pillar of fire for the camp, years after go-live. Skip it, or do it carelessly, and Murphy is waiting on the rocks.

Go liveAutopilot

Fifty-five days. Afterwards, the smallest event — a customs release registering as a twitch in the company's mainframe — shifts the item's ETA, estimates the port cost that hasn't yet been billed, and posts the corrected inventory value, all without a hand touching it. The audit catches billing errors no eye could have found — “and if it's only a hundred shekels, so what; in these volumes, would we ever have caught ten thousand?” And a question nobody could have answered before: who even knew what a KPI was — how a shipper performed, how the receiving ramp held up against standard times? The import leader now runs the whole thing by exception alone.

We got an autopilot — and we all speak one language now.

Which is, more or less, what Jenny had asked about. Yes. It can really happen.

If you will it, it is no dream.
— and fifty-five days later, it wasn't one.

With thanks to the import, logistics, finance and shipping teams — at a major vehicle importer whose name we keep, gratefully, on file — who lived this reform and let us tell it.