Lean before automation: what Valencia changed

I arrived in Valencia thinking I would learn tools for improving processes. What I actually learned was how to see waste I had been walking past for years.

In May and June 2026 I spent 38 hours in a classroom in Valencia, Spain, completing Green Belt training in Lean Office with SGS Productivity by Leansis. I went in expecting a toolkit. I came out with something more useful and slightly annoying: the ability to see waste I had been walking past for years.

Office waste hides better than factory waste

Lean grew up on a factory floor, where waste is physical. Inventory piles up. A machine sits idle. You can point at it.

In an office, the same waste exists and none of it is visible. Work in progress is an inbox. Overproduction is a report nobody reads. Waiting is a request sitting in a queue for three days. Rework is the third version of a file because the first two went to the wrong people. Motion is a person opening five systems to answer one question.

It is all there, and it is invisible, because it looks like people being busy. That is the trap. Busy is not the same as productive, and an office is very good at generating the appearance of the first.

The tools that actually earned their place

The training covered a lot — Lean culture, value versus waste, business strategy, balanced scorecards, team leadership, kaizen, A3, process mapping, SIPOC, value stream mapping, process reengineering, Kanban, 5S, PDCA, visual management, productivity and operational excellence. In practice, a few come out of the bag far more often than the rest.

• SIPOC, for settling the boundary. Most disagreements about a process are actually disagreements about where it starts and ends, and this surfaces that in twenty minutes.
• Value stream mapping, for the ratio. Separating processing time from elapsed time is consistently the moment a room goes quiet.
• A3, for discipline. One page forces you to state the problem, the current condition, the root cause and the countermeasure without hiding behind slides.
• PDCA, for honesty. It builds in the check. Improvement work without a check step is just change with good intentions.

Removing a step is cheaper, faster and more durable than automating it. It is also much harder to get approved, because it looks like less work.

Why this changed how I automate

Before the training I was already asking whether a process made sense. What I lacked was a rigorous way to demonstrate the answer to somebody else. Instinct is not persuasive in a governance meeting. A mapped value stream with the waiting time marked on it is.

That matters because the recommendation is often unwelcome. Telling a stakeholder that their automation request should instead be a redesign is a difficult conversation, and it needs evidence rather than opinion.

Standard work and the parts people skip

Two ideas from the training get dismissed in office environments and should not be. The first is 5S applied to digital work — not tidying for its own sake, but the observation that if five people store the same kind of file in five places under five naming conventions, everyone downstream pays for it forever, including any automation you later try to build on top.

The second is visual management. On a factory floor you can see the state of the work by standing in the room. In an office the state of the work is inside individual inboxes and systems, invisible to everyone including the people responsible for it. Making flow visible — a simple Kanban view of what is in progress, what is blocked and what is waiting on whom — regularly changes behaviour before anybody automates anything.

Standardisation carries a similar misunderstanding. People hear it as removing judgement. What it actually removes is unnecessary variation, so that the judgement can be spent where it matters. You cannot automate a process that is performed four different ways by four people, and you should not try — you should first find out which of the four is best and why the others differ.

The sequence

The order I work in now is simple enough to state in a sentence. Map the process as it really runs. Separate value from waste. Remove what should not exist. Standardise what remains. Then automate whatever is still there and still worth doing.

Skipping straight to the last step is the most common mistake in this field, and it is understandable, because it is the only step that produces something you can show people.

One caveat I try to be careful about: a Green Belt in Lean Office is a solid working foundation, not a claim to have mastered Six Sigma. The tools I use most are the ones above, applied to office and finance processes.