Setting Service Level Expectations (SLEs) in Kanban

Service Level Expectations (SLEs)

A Service Level Expectation (SLE) is a probabilistic forecast of how long items will take to move through your Kanban system. Unlike SLAs (which are contractual commitments), SLEs are forecasts based on historical data.

How SLEs Work

An SLE is expressed as a percentile of cycle time:

"85% of items are completed within 5 business days."

This means that historically, 85 out of 100 items finish within 5 days. The remaining 15% may take longer.

Setting Your First SLE

1. Collect cycle time data for at least 20-30 completed items.
2. Sort the cycle times from shortest to longest.
3. Find the 85th percentile value.

Example data (sorted cycle times in days):

1, 1, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 5, 5, 5, 5, 6, 7, 8, 12

20 items. The 85th percentile is the 17th value: 7 days.

Your SLE: "85% of items complete within 7 business days."

Using SLEs for Aging

Track the age of each in-progress item -- how long it has been since work started. Compare the age to your SLE:

• Green zone (age < 50% of SLE) -- On track.
• Yellow zone (age 50-85% of SLE) -- Needs attention.
• Red zone (age > 85% of SLE) -- At risk of breaching the SLE.

Use color labels to reflect these zones and review red-zone items in every standup.

Benefits of SLEs

• Predictable delivery -- Stakeholders know what to expect.
• Early warning system -- Aging items are flagged before they become overdue.
• Data-driven improvement -- If your SLE worsens over time, it signals a systemic problem.