Levels of service show up in most asset management frameworks as a required section, which is probably why so many of them read like they were written to satisfy a template rather than to describe what a utility is actually trying to deliver. The problem shows up the first time a board member, auditor, or funding reviewer asks a follow-up question, and the target on the page turns out to be aspirational rather than defensible.

The difference between a target and a level of service

"We aim to provide reliable water service" is a mission statement, not a level of service. A defensible level of service is specific enough to measure, tied to something your customers or regulators actually experience, and honest about where your system currently stands against it. "Fewer than four unplanned water outages per 1,000 connections per year, with average restoration time under three hours" is a level of service. It can be measured, tracked, and defended with data.

Where levels of service actually come from

The targets that hold up are rarely invented from scratch. They typically come from three places layered together:

  • Regulatory minimums, drinking water standards, permit conditions, and reporting requirements your utility already has to meet regardless of what you decide.
  • Customer expectations, which are often better understood through complaint patterns and service requests than through a formal survey, though a survey helps when you can do one.
  • What your system can actually deliver today, based on current condition and capacity, not an aspirational future state. A level of service your system can't currently meet isn't a target, it's a gap that belongs in your improvement plan, and the two should not be confused with each other.

Setting targets you can actually measure against

A level of service is only as useful as your ability to report against it consistently. Before finalizing a target, it's worth checking whether you can actually produce the number behind it today, from work orders, SCADA data, complaint logs, or inspection records, and if not, whether getting there is a reasonable near-term improvement or a multi-year data project. Setting a target you can't yet measure isn't wrong, but it should be labeled as a target you're building toward, not one you're currently meeting.

Connecting levels of service to risk and capital planning

Levels of service aren't meant to sit on their own. They're the standard against which your risk assessment and capital plan get justified. When a capital project shows up in a CIP, the strongest justification connects it directly to a level of service gap: this lift station has failed its response-time target three times in two years, and this project closes that gap. That link is what turns a capital plan from a list of requests into a document a board can actually evaluate on its merits.

Revisiting targets honestly

Levels of service aren't permanent. As your system's condition improves, or as customer expectations and regulatory requirements shift, the targets should be revisited, and where they haven't been met, that gap should be reported honestly rather than quietly adjusted downward to make the numbers look better. A level of service that's been quietly softened every year it's missed stops being useful as a planning tool long before anyone says so out loud.