Who Runs It Afterwards

The project ends at go-live. The system then generates work every week, indefinitely, and that work is routinely unassigned.

The recurring work

Corrections: approving them, chasing the ones that stall.

Refusal investigation: which site, which cause, what to fix.

Radii: adjusting after building work, new sites, changed entrances.

Devices and accounts: joiners, leavers, replaced phones.

Platform updates: the check after every operating system and vendor release.

Retention: verifying deletion ran.

Requests: subject access, dispute evidence.

Totalled, this is a part-time role, usually absorbed by whoever runs payroll and was not asked.

What goes wrong without an owner

Refusal rates climb and nobody notices, because nobody is looking at the number.

Corrections pile up until payroll, which is when they become pay errors.

A radius is never adjusted after an entrance moves, so one site refuses everything.

Features arrive enabled in an update and nobody checks.

Retention never runs, so coordinates accumulate indefinitely.

None of these announces itself. They surface in a grievance.

Naming the owner

One person, named, with allocated time.

With authority over configuration, or every radius change becomes a request to someone busier.

With a deputy, because corrections do not pause for annual leave.

And with the standing to decline the requests in the refusal note.

The runbook

How to approve corrections, and the deadline.

How to investigate a refusal: site first, coverage second, person last.

How to measure a site and set a radius.

The post-release check.

Who authorises access to coordinates and how it is recorded.

Written so the deputy can follow it, which is the test.

What to report

Monthly: refusal rate by site, correction volume and turnaround, fallback usage.

Quarterly: accuracy distribution, access review, deletion verification.

Annually: did the measures move, is the collection still what was agreed, would you deploy it the same way again.

To the workforce as well as to management, because publishing the refusal rate is what signals that failures are the system's problem rather than theirs.

Publish the refusal rate

To the workforce, not only to management.

It signals that failures belong to the system rather than to them.

It invites people to report a site that is failing, which is how you find problems early.

And it commits you to fixing what it shows, which is the point.

Very few deployments do this and it is the cheapest trust-building measure available.

A concrete product reference

When translating this principle into a buying test, this workflow reference provides a concrete workflow reference. Verify current behaviour in a trial and judge it against the purpose and limits above.