Anyone can copy a price list. This is the part they cannot copy: what happened after the price was agreed, published whether or not it flatters us.
Reporting on September 2026, computed Oct 1, 2026
We have started recording, but no figure below yet rests on enough completed work to be read as a rate. Each one shows its sample size so you can see exactly how thin it is. We would rather show you that than wait until the numbers flatter us.
Nothing to report
Measured over 0 projects started in the period
How it is measured: Of the projects started in the period, the share whose proposal cited a tier that resolves against the published price list. A proposal that cited no tier counts as a miss, not as an exclusion.
As of Oct 1, 2026
Nothing to report
Measured over 0 projects completed in the period
How it is measured: Of the projects completed in the period, the share completed on or before the date the proposal promised, not a revised target. Projects whose proposal named no date leave both sides of the ratio, and the count is stated.
As of Oct 1, 2026
0%
Measured over 1 projects active in the period
Small sample of 1. Read this as an indication, not a rate.
How it is measured: Of the projects active in the period, the share with at least one change order issued. Declined and withdrawn change orders count: the question is how often the agreed scope turned out to be wrong, not how often we were paid more.
As of Oct 1, 2026
Nothing to report
Measured over 0 change orders issued in the period
How it is measured: The mean value of every change order issued in the period, including declined and withdrawn ones. Averaging only approved change orders would report the ones a client accepted and hide the ones they refused.
As of Oct 1, 2026
Nothing to report
Measured over 0 checks run in the period
How it is measured: Successful checks as a share of every check run against monitored client sites in the period. All of them count, including checks that failed during a deployment we caused.
As of Oct 1, 2026
Nothing to report
Measured over 0 restore tests due in the period
How it is measured: Of the restore tests due in the period, the share completed by their due date. A test that was escalated and then completed late counts as a miss.
As of Oct 1, 2026
Nothing to report
Measured over 0 client messages answered in the period
How it is measured: The median time from a client message to our first reply, broken down by retainer tier. Measured from every message that opened a thread, including ones sent outside working hours.
As of Oct 1, 2026
Every figure is computed from our own records for one calendar month and refreshed monthly. Each publishes its sample size, and any figure measured over fewer than 5 records is marked as a small sample rather than withheld.
A bad month is published like any other. A record that only appears when it is good is advertising, and would be worth nothing as evidence.
If you are a client and a figure here does not match your experience of working with us, tell us. The underlying records are retained, and we will correct the page rather than argue with you about it.