Analytics · Sample size

Forty cars can tell
a loud story.

A small property produces real data, but not much of it. One stalled elevator, one dead battery, or one ten-car departure wave can swing a nightly average enough to trigger a staffing change the next night reverses.

The discipline

Treat a single night as an observation, not a verdict.

Small samples are useful for finding incidents and asking questions. They are weak foundations for declaring a stable improvement or decline, especially when the mix of guests and conditions changes with each shift.

Pool several comparable shifts, report the count, preserve the distribution, and annotate known disruptions. Do not average away the story of the individual nights.

Illustrative swing

A few cars can move the whole headline

01

See the leverage of the tail

This example is illustrative and uses invented values.

40
illustrative completed retrievals
4:30
illustrative mean before three delays
+3
illustrative unusually slow retrievals
5:25
illustrative mean after those delays

Three observations changed the headline by nearly a minute. That may expose a real failure, but it does not establish a new normal.

02

Separate noise from an incident worth fixing

Repeat

Did the pattern recur?

One bad elevator trip is an incident. Repeated delay on the same floor is an operating constraint.

Cluster

Did the delays share a cause?

Open the session records. Similar timestamps, zones, handoffs, or vehicle states can turn outliers into one mechanism.

Compare

Was demand actually similar?

A 40-car restaurant night and a 40-car event release do not provide the same service test.

Count

How many observations support the claim?

Always show the count next to the metric so a viewer can distinguish a night from a season.

03

Build a rolling view without hiding change

Single shift
sensitive
Comparable rolling group
steadier
All-time average
too slow

A rolling group of comparable shifts balances responsiveness and stability. An all-time average can become so stable that it ignores the current operation.

04

Use a decision threshold, not a mood

Before seeing the next number, define what evidence would trigger action.

Immediate

Safety and severe failures

Some incidents deserve action after one occurrence. Sample-size caution is not an excuse to wait.

Monitor

Small service movement

Require repetition across comparable shifts before changing staffing or process.

Investigate

Large unexplained swing

Open the underlying records before labeling the whole team or property.

The reporting rule

Show the dot before the line.

Trend lines are useful, but a manager should still be able to see the individual shifts and counts beneath them. That preserves both the pattern and the night that broke it.

The takeaway: small data is not useless. It simply demands smaller claims, comparable pooling, and faster access to the records underneath.

Keep reading.

AnalyticsResearch

Which Change Actually Caused the Improvement?

When an operation changes staffing, staging, signage, and request flow at once, the before-and-after story cannot identify a winner. A staggered rollout creates a more credible comparison.

May 12, 2026 · 8 min readRead
OperationsResearch

The Anatomy of an Efficient Valet

Most valet operations are built to stall - rarely because of the team, almost always because of the layout. Eight principles from how the best operations actually move cars.

Jun 26, 2026 · 7 min readRead
AnalyticsKpi

Why Retrieval Time Is the KPI That Matters Most

If you could track exactly one number in your valet operation, this is it. Retrieval time is the rare metric that's a service score, a financial signal, and a management tool at once.

Jun 16, 2026 · 5 min readRead

When reading is not enough

See it on your drive.

Twenty minutes on your own property, with your own volumes. We would rather show you the parts an article can only describe.

or keep reading the journal