← FleetLedger blog

The fleet manager left, and the spreadsheet that held the whole policy together left with them

The operations director signed off on it without ever saying yes to it. The fleet manager kept a spreadsheet that matched vehicles to policies, drivers to vehicles, and registration numbers to the cover the broker thought was in force. For four years nobody asked where the reconciliation lived, because it lived in fleet_master_v7.xlsx and in one person's head. The assumption was that the process was a system. It was a person.

Then the fleet manager resigned. The handover was tidy, but what did not transfer was the part that mattered: why three vehicles still on the schedule had been sold, why two leases added in March were not on it, and which of the APK dates in column F were real versus copied forward from last year. At the next renewal, nine vehicles had a status nobody could confirm.

The single-spreadsheet failure mode and why it's so common in fleets

A single point of failure is any component a system depends on with no backup behind it. In most fleet programs that component is a person and the file they own. Fleets drift faster than other insured asset classes because the list changes every week: a van sold at auction, a lease swapped, a pool car reassigned to a new depot. When one human keeping one sheet has to catch each change and record it in the one place that feeds the broker, the control and the key-person risk are the same object.

It works right up until it doesn't, because the reconciliation is accurate as long as someone remembers the exceptions, not because the exceptions are recorded. Pull that person out and you don't lose a file. You lose the index that tells you which cells to trust.

What breaks first when the keeper of the list leaves: renewals, endorsements, claims

Renewal breaks first because it is the first hard deadline after departure. The insurer prices off the schedule you confirm, and if you cannot confirm it, you either over-declare and overpay or under-declare and carry a gap you can't size. Endorsements break next: a vehicle added mid-term needs a cover note and a schedule update, and with no owner of the list, additions stop happening cleanly. The claim is the one that actually costs you, when the first incident after the fleet manager leaves turns on whether a specific registration was on cover that day, with an adjuster asking for the schedule the departed employee used to maintain.

Why a system of record is a continuity control, not a software upgrade

The instinct after this happens is to buy a tool, and that framing is part of the problem. A system of record is not a nicer spreadsheet; it is the thing that lets the reconciliation survive a resignation. The test is simple: if the person who maintains the fleet list disappears tomorrow, can the next person produce an accurate schedule and bordereaux without reconstructing four years of someone's memory?

The mechanics matter. Most bordereaux are still re-keyed by hand from policy systems into a sheet before they reach the insurer, which is exactly where errors and lateness enter. Hold vehicles, drivers, and cover as records with status and history, and "is this car on the policy today" has an auditable answer instead of an institutional rumor.

Offboarding a fleet manager without losing the reconciliation

Treat the reconciliation as an asset you are repossessing, not a favor the leaver does in their last fortnight. Before the notice period ends, get these out of one head and into the record:

The schedule on the day they leave should be one a stranger could defend to an underwriter.

FleetLedger exists so the list is not a person. Each fleet program runs as its own record of vehicles, drivers, and cover, reconciled against registration data continuously, so the schedule and bordereaux stay current whether or not the person who owned the spreadsheet is still in the building. A resignation should cost you a colleague, not your system of record. See how FleetLedger holds the reconciliation when people change.