liveBook · Manning

Waste

As explained in Automating API Delivery

Waste is process effort or elapsed time that does not contribute value to delivering the intended product or change.

An API change can take weeks to reach approval even when the work itself takes hours. Waste is effort or elapsed time in that delivery process that does not create customer value. Finding it matters before adding automation: a faster individual step may not improve the end-to-end workflow. section 2.0 section 2.4.4

Lean thinking starts by asking what the customer finds useful, then tracing the activities that deliver it. Analysis, design, coding, and API documentation can be value-adding because they produce outputs valuable to the customer. The process also contains value-enabling work and non-value-adding work; do not assume every activity that fails to transform an input is waste. section 2.11

To locate waste in an API workflow:

  • Map the steps in the value stream and distinguish value-adding, value-enabling, and non-value-adding activity. section 2.0 section 2.11
  • Compare time spent doing work with total elapsed time. Look for queues and periods when a change waits for someone to resume it. section 2.11 section 2.4.4
  • Use the map to choose an improvement, then check whether it improves delivery of valuable, high-quality APIs—not merely the speed of one automated task. section 2.0 section 2.4.4

Approval time versus working time

Adding one API operation takes 16 working days (128 hours) of elapsed time but 5.25 hours of actual work. The activity ratio is 5.25 ÷ 128, or 4.1%; for the remaining 95.9%, the change sits idle. A review amendment and final approval take an hour of work yet span a working day because engineers are occupied with other tasks. The map makes that waiting visible as a target for improvement. section 2.4.4

How Automating API Delivery covers it

How deep it goes
Classifies production activities and quantifies idle time with a worked activity-ratio calculation; does not specify an automated waste-detection tool.
Examples
Uses approval of a new API operation: 5.25 hours of work across 128 hours of elapsed time.
What you can do after
Diagnose/debug with a process map and activity ratio to find waiting that dominates an API change’s elapsed time.
How it connects
Connects lean thinking, value stream mapping, queues, and APIOps process improvement.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Automating API Delivery. The section links point to the parts of the book it draws on.