Ohno's Seven Wastes, Translated to the Office

An engineering services firm, 240 people. A commercial proposal goes through costing, technical review, margin approval, and an executive signature. Along the way it gets built in a spreadsheet, pasted into a deck, exported to PDF, attached to an email, discussed in a chat group, and sent back with a change request that means redoing the spreadsheet.

Nobody would call that waste. Everyone worked, everyone replied, the client got the proposal. Except that of the 14 hours of human work the proposal consumed, somewhere between 5 and 7 added nothing the client would pay for.

01The idea: what the customer doesn't pay for #

Taiichi Ohno spent decades at Toyota building the production system that carries the company's name, and published his synthesis in 1978. The central concept is muda: any activity that consumes resources and creates no value for whoever receives the result. Ohno catalogued seven categories — overproduction, waiting, transportation, over-processing, inventory, motion, and defects.

He wasn't proposing that people work harder. The argument is the opposite: working harder inside a process full of muda increases effort, not results. Ohno also insisted waste is rarely eliminated by decree — it becomes visible when you observe the flow up close, where the work actually happens.

02The seven, translated #

1. Overproduction — the weekly report seven people receive and two open. Producing information before a decision exists that depends on it is the most expensive waste, because it generates all the others.
2. Waiting — the proposal sitting three days waiting on margin approval. It isn't the person who's idle: it's the work.
3. Transportation — the same information moving between spreadsheet, email, CRM, and deck. Every crossing costs time and creates a chance for divergence.
4. Over-processing — the field filled in because the form demands it, the slide formatted for a meeting that runs eight minutes, the double check that exists because someone made a mistake once.
5. Inventory — 40 open tickets nobody will look at this week, 12 projects started and stalled. Work inventory doesn't spoil on the shelf, but it ages: by the time someone picks it up, the context is gone.
6. Motion — the hunt for the right file, the question 'who has the latest version?', the fifteen minutes spent figuring out which tool that decision was recorded in.
7. Defects — approval rework: the document that comes back because it was missing a piece of data the requester never knew was needed.

Womack and Jones later added an eighth item Ohno didn't list: the waste of not using what people know. In a mid-size company it's the most common of all — the person who has run the process for four years knows exactly where it breaks and has never been asked.

03Why the office hides it better #

On the factory floor, inventory is physical: it takes up space, blocks the aisle, shows up in the photo. In administrative work, inventory is a list of unanswered emails and waiting is an invisible item on somebody's calendar.

Add a cultural detail: in the office, being busy signals value. Someone who spends the day answering messages about task status looks productive, even though nearly all of that activity is information transport — pure waste, in Ohno's classification.

And since each area uses whichever tool it prefers, transport between systems became part of the job rather than an anomaly. It's the kind of cost nobody approves in a budget and everybody pays every day.

04The cost of approval rework #

At the scale of a 240-person mid-size Brazilian company (figures in Brazilian reais), with 60 proposals a month. The assumptions are conservative estimates, stated so you can redo the math with your own numbers:

Want to see how this looks inside a real operation? Explore the platform.

The math: 60 proposals × 1.8 round trips × 2.5 hours = 270 hours a month; at R$ 90 an hour, R$ 24.3 mil a month. And that's only the defect category — it doesn't count waiting, transport, or report overproduction. None of those hours show up as a cost anywhere: they're buried inside salaries the company would pay regardless, which is exactly why nobody sees them.

05Where this gets stuck in practice #

• The waste is distributed. No area has an obvious 5 hours to cut; each one has 40 minutes. Only a view of the whole flow makes the total visible.
• Whoever sees it doesn't decide, whoever decides doesn't see it. The person feeling the rework sits three levels below whoever could change the approval rule.
• Cutting a step looks like cutting control. The double review was born from a real mistake. Removing it means replacing control with criteria — and criteria require data.

06What data-driven management answers #

Administrative waste is invisible because it isn't recorded. When every step of the flow has an entry, an owner, an exit, and a timestamp, the seven categories stop being a conceptual exercise: waiting becomes queue time, defects become a send-back count, transport becomes the number of systems crossed per document.

One honest caveat: reducing waste outside the point that limits the operation doesn't change the lead time the customer feels — that's Goldratt's argument about the bottleneck. And resist the urge to solve everything with a bot before simplifying; automating a bad process only speeds up the error.

Bringing the work, the process, and the record into one place is exactly what the Collaborative Work Management category describes, and what the processes and automations modules in the platform do in practice. To see the design by area before deciding where to start, take a look at the solutions by operation.

Ohno wasn't asking for more effort. He was asking us to stop paying for work the customer will never receive.