Our service delivery cycle is lumpy, which causes our weekly Scorecard numbers to swing wildly from green to red week over week. Should we track rolling four-week averages on our Scorecard or stick to rigid weekly snapshots?

Question

Tyler Smith · Accepted Answer

Stick to weekly snapshots. Smoothing data with rolling four-week averages is tempting for businesses with lumpy delivery or uneven sales cycles, but rolling averages hide urgent operational problems. They dilute the impact of a terrible week and delay your leadership team from identifying root causes in your Level 10 Meeting™.

When you use a rolling average, a disastrous zero-production week is masked by three previous strong weeks. By the time the rolling average turns red, three to four weeks of damage have already compounded. The purpose of an EOS® Scorecard is not to make the leadership team feel comfortable with smooth visual trends. The purpose is to provide an immediate, unfiltered weekly pulse that catches operational decay early.

If your operational reality is genuinely lumpy, adjust how you evaluate the data rather than altering the math:
- Keep the metric as a pure weekly number. Record the exact activity or output generated over the past seven days.
- Establish an intelligent target range or trigger rule. If you know project sign-offs naturally cluster bi-weekly, set an expectation that two consecutive off-target weeks triggers an automatic drop to the Issues List for IDS®.
- Pair the lumpy output metric with a leading pipeline metric. If completed client handoffs swing wildly, track weekly milestones ready for client inspection. That upstream number will show whether the downstream lumpiness is healthy workflow cadence or erratic execution.

Run on raw weekly data. Living by true numbers requires confronting operational variance directly every seven days.