Your team isn’t slow. It’s leaking time.
Eight ways work leaks time — Lean calls them DOWNTIME. Once you can name them, you can’t stop seeing them.
Your team works hard. The hours are long, the standups are full and everyone looks busy — and yet the thing that should take a day takes three. Velocity feels slower than the effort warrants. Retros surface the same complaints again and again yet nothing moves. If you’ve started to suspect the team is just slow, stop.
It isn’t. The time isn’t evaporating. It’s leaking — escaping through gaps you walk past every day without seeing them. And here’s the part that matters: the gaps aren’t a mystery. There are exactly eight of them, and they were named decades ago, on the factory floors of postwar Japan.
When Dave Brailsford ran British Cycling and then founded Team Sky, he didn’t chase one breakthrough. He hunted one-percent improvements everywhere — the bikes, the riders’ sleep, even the way they washed their hands to dodge a cold before a race. Stack enough of them, the theory went, and a country that had never produced a Tour de France winner takes five of the next six. That’s the arithmetic James Clear made famous: get one percent better every day and you compound to thirty-seven times better in just one year. The deeper idea is older — kaizen, the culture of relentless small improvements that Toyota turned into an industrial empire, one daily fix at a time.12
The catch is that improvement is the work everyone defers. It’s the first thing cut when a deadline looms and the last thing scheduled when one doesn’t — a luxury we’ll get to once the “real” work is handled. That’s exactly backwards. Finding the next one percent is what makes next week cheaper than this one. And the fastest place to find it is wherever your time is leaking.
If you’re new, welcome to Customer Obsessed Engineering! I publish about one article each week. Free subscribers can read about half of every article, plus all of my free articles.
So before you can plug the leaks, you have to see them. Here they are — the eight ways work leaks time, the list Toyota’s engineers assembled and the rest of the world adopted. Lean remembers it with a mnemonic you won’t forget once you’ve heard it: DOWNTIME. And yes, it applies to software.3
Most teams already have words for some of this. They call it “process overhead” or “tech debt” or “scope creep” — phrases that stay vague on purpose, so everyone can nod without anyone having to fix anything. DOWNTIME is the vocabulary those euphemisms exist to replace. You can’t kill what you can’t name. Eight letters, eight leaks: defect, overproduction, waiting, non-utilized talent, transportation, inventory, motion and extra processing. Read that once and you’ll start seeing it everywhere — which is the entire point.
The ones you know too well
You won’t feel all eight equally. Most teams are dominated by two or three. Here are the ones that tend to draw blood.
Waiting. The pull request that sat for two days while its author moved on to something else. The deploy you watched spin while you refreshed a dashboard and pretended that counted as work. Nobody did anything wrong — which is exactly why waiting is invisible. The work just sat there, and so did you.
Defect — the rework tax. You build the feature and you build it well. Then the acceptance criteria turn out to have been wrong and you build it again. In software, rework is often the most expensive waste you carry, because it’s work you already paid for, now billed a second time. Every “wait, that’s not what we meant” is a withdrawal from an account you didn’t know was still open.4
Motion — death by a thousand switches. Context-switching across tools, the same manual clicks twenty times a day, the hand-rolled deploy sequence you’ve run a hundred times. The cost isn’t the click — it’s the recovery. Gloria Mark’s research clocked the average return from an interruption at twenty-three minutes. Get pulled away every few minutes and you never actually arrive at the work at all.5
Extra processing — the meeting that decided nothing. The ceremony that produces no decision. The low-risk change you tested five ways. The elegant abstraction built for a second use case that never showed up. All of it is real effort. None of it is progress. None of it adds real value.
The quieter four
The other four leak just as quietly — but they leak all the same.
Overproduction — building ahead of demand. The feature shipped before anyone asked for it, the configuration knobs added “just in case,” the speculative endpoints built for a use case that never arrives. It looks like output. It’s effort that earns nothing until someone needs it — and most of it, nobody ever does.
Transportation — the handoff tax. Every time work crosses a border — dev to QA, one team to another, a ticket bounced to a different queue — it stops and waits for someone on the far side to pick it up. The work itself doesn’t change. It just accrues delay at each frontier, like luggage transferring between flights. (This is the cousin of motion: motion is you moving between tools, transportation is the work moving between people.)
Inventory — the column that only grows. Half-finished features, branches waiting to merge, tickets piling up in a “ready for review” lane nobody clears. Unlike a warehouse, software inventory is invisible, so it accumulates without anyone flinching. Every item is capital you’ve already spent and haven’t earned back.
Non-utilized talent — the most expensive leak of all. Your strongest engineer triaging tickets a junior could clear. The designer who never designs because she’s stuck in status meetings. The architect rubber-stamping config changes. It’s the waste that doesn’t just cost hours — it’s the one that quietly walks people out the door.
Notice what just happened as you read those. You’ve been losing time to them for years. The instant they have names, they stop being the weather — just how things are around here — and become a list of things you can go after. That is the whole move. Naming waste is how you stop tolerating it.
This newsletter grows by word of mouth… I’d really, truly appreciate it if you could refer a friend. Your referrals make it worthwhile.
Stop guessing, start measuring
Want to know which of the eight is actually costing you the most? Don’t argue about it in a retro — measure it. For one sprint, have everyone jot down where their time went and whether it moved the work forward. Two minutes a day. Add it up and look at the shape of it. That’s a waste walk, and it’s about as complicated as it sounds.6
What comes back will reorder your priorities on the spot. The bottleneck you were certain about turns out to be a rounding error. The “quick” thing nobody ever questioned turns out to be the biggest line on the page. The vague dread that “we’re just slow” snaps into focus as something you can actually attack — this many hours bled to rework, this many to manual steps a script should have saved us. You come in with a hunch. You walk out with a map that has an “X” on it and an action list.
What you actually get back
The obvious prize is the hours. But the hours aren’t what makes this worth doing. Look at the eight wastes again and notice what they share: every one of them is a part of the job you already hate. Rework, waiting, manual drudgery, meetings that end in nothing — that’s the misery. It isn’t the work you signed up for. It’s the sludge that collects around it.
So when you cut the waste, you’re not just going faster. You’re scraping away the worst parts of the week and leaving the work you actually wanted to do — the building, the solving, the craft. A team that kills its waste doesn’t just ship more. It spends its days on better things and gets to go home at a decent hour.
And you trade in something subtler: the shrug. “That’s just how it is here” is the most expensive sentence in software, because it ends the conversation before it starts. Naming and measuring waste ends the shrug instead. It hands the team back the knowledge that “the way things are” is a choice and not a fact — that the team gets a vote.
That’s the one percent. Not a heroic, quarter-long transformation — one leak found and plugged this sprint, then the next one. It’s how Brailsford’s riders went from also-rans to the top of the podium: not a single breakthrough, a hundred small ones. Your team can run the same play, and it starts with a single word for the thing that’s been slowing you down.
Try this tomorrow
You don’t need a platform, a budget or anyone’s permission to start. You need eight words and a little attention.
Tomorrow, just name things as they happen. The bug you’re fixing because the criteria were vague — defect. The forty minutes lost waiting on a deploy — waiting. The fourth manual step in a release you’ve run a hundred times — motion. Don’t fix anything yet. Just see it. By Friday the list will have stopped being a mnemonic and started being a habit — the habit of a team that’s about to get one percent better, and then keep going.
When that habit sticks, take it to the next level: run a waste walk. Give it a single sprint and you stop guessing entirely — the data names your biggest leak for you and hands you the business case to fix the ones that need leadership to move. I’ve written the whole playbook for it in Stop doing things that don’t make us faster, better, more focused.
Word of mouth is how this newsletter grows, and your support means everything.
Dave Brailsford’s “aggregation of marginal gains” at British Cycling and Team Sky. See the World Economic Forum, This coach improved everything by 1%, 2016, and James Clear, Marginal Gains.
The “one percent a day → 37x a year” framing is James Clear’s; the arithmetic is plain compounding (1.01³⁶⁵ ≈ 37.8). See James Clear, Continuous Improvement. Kaizen itself is the continuous-improvement philosophy at the heart of the Toyota Production System — a culture of small daily fixes rather than a literal formula. See Masaaki Imai, Gemba Kaizen (McGraw-Hill, 1997).
The seven original wastes (muda) come out of the Toyota Production System, codified by its architects Taiichi Ohno and Shigeo Shingo; modern Lean adds an eighth, non-utilized talent, for the DOWNTIME mnemonic. See Taiichi Ohno, Toyota Production System (Productivity Press, 1988), and the Lean Enterprise Institute, Muda, mura, muri. The software-specific examples here are drawn from the GembaKai waste walk facilitation guide.
That rework is the costliest waste is not just intuition. Barry Boehm and Victor Basili found that fixing a defect after delivery is often 100 times more expensive than catching it during requirements or design, and that software projects spend roughly 40–50% of their effort on avoidable rework. See B. Boehm and V. Basili, Software Defect Reduction Top 10 List, IEEE Computer, January 2001.
The widely cited ~23-minute recovery figure (23 min 15 sec) comes from Gloria Mark, Victor González and Justin Harris, No Task Left Behind? Examining the Nature of Fragmented Work (CHI 2005). Her later study, The Cost of Interrupted Work: More Speed and Stress (CHI 2008), found that interrupted work is finished faster but at the cost of higher stress and effort. For more on this particular leak, see Context switching is killing your gains (part 2).
For the full mechanics — running the exercise, reading the data and building the business case — see Stop doing things that don’t make us faster, better, more focused. The underlying technique of sorting effort into value-adding and non-value-adding work is value stream mapping; see Value stream engineering in a nutshell. For the platform that automates it, check out GembaKai.