Certain key activities in our workflow happen in irregular batches rather than smooth weekly flows, causing that Scorecard number to swing between red and green every other week. How do we track batched or lumpy operational data without generating false alarm IDS discussions every Tuesday?

Question

Tyler Smith · Accepted Answer

When an operational activity happens in natural batches, applying a rigid weekly flat target creates artificial drama. Your Level 10 Meeting™ agenda gets cluttered with false alarm misses on week one, followed by meaningless celebration on week two.

You have three practical ways to stabilize lumpy metrics on an EOS Scorecard:
- Use a rolling four-week average: Instead of tracking single-week output, track the trailing four-week sum divided by four. This smooths out seasonal or bi-weekly batch processing while still showing whether the overall trajectory is declining.
- Measure the leading staging activity instead of the batched release: If your development team deploys client software every three weeks, do not track deployments. Track weekly pull requests reviewed or milestone validations completed. Measure the daily work that feeds the batch, not the release event.
- Measure cycle time or queue aging: If invoices are sent out bi-weekly, do not measure invoices sent per week. Measure unbilled work-in-progress days or days from deliverable completion to invoice generation.

Never accept the excuse that an operational department cannot be measured simply because their output is lumpy. Shift the measurable from the final batched release to the continuous flow of work feeding that release.