The winter operations KPIs actually worth tracking

After a storm, most operations do a gut-feel retrospective. "That one went okay." "That one was a mess." The instinct is right — you should be reviewing storms — but gut feel does not tell you which lever to pull next time. A short set of the right metrics does. The trick is choosing metrics that measure the outcome customers actually care about, not the ones that are easy to count.

The vanity metrics to distrust

Two numbers are tempting and misleading. Total snow moved and total hours worked both feel like productivity, but neither tells you whether you met your obligations. You can move a lot of snow and still blow every premium SLA; you can log a lot of hours and still have machines idling near each other while a property waits. Volume is an input, not an outcome.

The metrics worth tracking

1. SLA attainment

The single most important number: what fraction of properties were serviced within their contracted window. This is the outcome the customer is buying and the one liability turns on. Track it per storm and, ideally, per account tier, because a 95% overall attainment can hide a disaster on your premium accounts. If you measure one thing, measure this.

2. Coverage over time

SLA attainment is a final score; coverage is the play-by-play. What share of your properties were complete at each point through the storm? A coverage curve that climbs steadily is a healthy storm; one that stalls tells you where the fleet got stuck. Watching it live also lets you intervene before an SLA is actually missed.

3. Idle and overlap time

This is the efficiency metric fixed routes hide. How much time did machines spend idle or working redundantly close to a peer while other properties waited? In a dense operation this is where coverage is quietly lost, and it is the slack a dynamic dispatcher is designed to reclaim. Falling idle-and-overlap time is the clearest sign your dispatch is getting tighter.

4. Plow-lag and recheck rate

Covered in depth elsewhere, but it belongs on the list: how often properties needed a return trip and how long re-accumulation took. It is the hidden labor cost of the storm and it points directly at accounts that need re-timing or re-pricing.

5. Machines-to-coverage ratio

How many properties did you cover on time per machine deployed? This is the number that justifies your fleet size and reveals whether adding a machine actually bought coverage or just added cost. It is also the number a good simulation lets you test before the season instead of discovering after it.

Measure the outcome the customer buys — properties serviced within the window — not the volume that is merely easy to count.

The point of measuring

None of this matters if the numbers just sit in a report. The reason to track SLA attainment, coverage, idle time, plow-lag, and machines-to-coverage is that each one points at a specific decision: staff differently, re-time a property, re-price an account, add or drop a machine, tighten the dispatch. A storm retrospective built on these five turns "that one felt bad" into "premium attainment dropped because two machines overlapped on the north cluster while the south waited — here is the fix." That is the difference between reviewing storms and improving them.

How Snowmass tracks them

None of this works if measuring the five is its own project. So Snowmass computes them from data the storm already produces — no extra logging, no spreadsheet after the fact. Your dashboard carries a storm-performance panel: SLA attainment, coverage, idle-and-overlap, recheck rate, and properties-per-machine, shown both season-to-date and for your latest storm, with the coverage curve drawn as it built through that storm. You set one contracted window — the hours a property should be cleared within — and attainment measures against it; leave it unset and the metric simply stays hidden rather than inventing a target you never agreed to.

The retrospective question — are we getting better? — is a report of its own. Pick any date range and Snowmass charts each KPI storm over storm, so a season of tightening dispatch shows up as a line that climbs (attainment, coverage) or falls (idle, overlap, plow-lag). That is the difference between a number on a dashboard and a decision: you can see the north-cluster overlap shrink after you re-timed it, and prove the machine you added actually bought coverage.

Key takeaways

• Distrust volume metrics — snow moved and hours worked are inputs, not outcomes.
• Track five that matter: SLA attainment, coverage over time, idle/overlap time, plow-lag/recheck rate, and machines-to-coverage.
• Measure SLA attainment per account tier — a good overall number can hide a premium-account failure.
• Each metric should point at a decision (staff, re-time, re-price, add a machine, tighten dispatch) or it is just a report.