Measured properly, or not claimed at all.

Before the system answers a single call, it listens for two weeks. Contacts are logged, outcomes are recorded, and nothing is changed. Without that fortnight, the word before means whatever a spreadsheet says it means, and every improvement claim afterwards rests on a number nobody wrote down at the time. Two weeks is the shortest window that survives a bad Monday and a quiet public holiday.

A share of contacts is routed exactly as it was before and never touched by the system. It runs in the same weeks, against the same demand, with the same staff. That is the only way to tell whether a lift came from the build or from the season. Where volume is too low for a holdout to mean anything, we say so and agree a different defensible comparison before launch rather than pretending afterwards.

Show rate, rebook rate, recovered contacts, escalation rate, average handling time, utilization by hour and by practitioner. All of them already exist in your business, which is the point. A metric invented for the report is a metric designed to pass.

Before the build starts, we agree the single number the project is judged by. It is written into the scope. Deciding it afterwards is how vendors end up reporting whichever chart looks best, and how clients end up unable to say whether the thing worked.

We will not put an invented percentage on a landing page. There is no plus forty-two percent anywhere on this site, and there will not be one until it comes out of a client's own reporting with their permission to publish it. Anyone can print a number. The question is whether it survives someone checking it.