Value Stream

In 1950, a Toyota engineer named Taiichi Ohno walked the factory floor and asked a question that would reshape global manufacturing: where does the time actually go? Not the machining time. Not the assembly time. The total elapsed time from raw steel entering the plant to a finished car leaving the dock. His answer was devastating. In most manufacturing processes, the material was being worked on — cut, stamped, welded, painted — for less than 5% of the total time it spent in the system. The other 95% was waiting. Waiting in inventory buffers. Waiting for the next workstation. Waiting for an inspection. Waiting for a signature. The car spent weeks in the factory, and the factory touched it for hours. Ohno called this the value stream — the complete sequence of steps, from raw material to delivered product, that a process performs. His insight was not that factories should work faster. His insight was that they should wait less.

A value stream map is the diagnostic tool. You draw every step in the process, from first input to final output, and for each step you record two numbers: processing time (the time the step actually adds value) and lead time (the total elapsed time including all waiting). The ratio between these two numbers is the process efficiency. In most organisations, across most processes, the ratio is horrifying. A mortgage application takes 45 days. Actual processing time: 3 hours. A software feature takes 6 weeks from spec to production. Actual development time: 4 days. A patient referral takes 3 months. Actual clinical work: 90 minutes. The value stream map exposes the gap between the work and the waiting — and the gap is almost always larger than anyone expects.

The Toyota Production System built an entire manufacturing philosophy on this observation. If 95% of elapsed time is waste — waiting, transport, rework, overproduction, unnecessary motion — then the highest-leverage improvement is not making the 5% faster. It is eliminating the 95%. Pull systems replace push systems. Single-piece flow replaces batch processing. Kanban boards make work-in-progress visible. The result: Toyota's production lead time dropped from weeks to hours while quality improved. The paradox — faster and better simultaneously — made sense only through the value stream lens. Toyota was not speeding up the work. Toyota was eliminating the waiting between the work.

The software industry rediscovered this principle sixty years later and called it DevOps. A typical enterprise software team in 2010 wrote code in two-week sprints, then waited three weeks for code review, two weeks for QA, one week for staging, and two weeks for change approval before reaching production. Total elapsed time: ten weeks. Actual development time: two weeks. The DevOps movement's core insight was pure Ohno: optimise the value stream, not individual steps. Continuous integration eliminated the code review queue. Automated testing eliminated the QA queue. Infrastructure as code eliminated the staging queue. Feature flags eliminated the deployment approval queue. Companies that adopted DevOps practices reduced deployment lead times from months to minutes — not by writing code faster but by eliminating the waiting that surrounded the code.