Why it matters
What this principle is for.
A high hiring bar is only the start. Today's standard will not be enough tomorrow. Everyone is expected to raise their own bar. Customers will not lower theirs.
Quality is how you keep trust. When a product, a service, or a process fails, confidence drops in the thing and in the people who shipped it. That is true whether the consumer is an end customer or another team.
This principle shows up as operational excellence. You never assume the current process is good enough. You pick the few metrics that tell you the truth. You fix the root cause so the defect cannot travel. You balance this with Bias for Action. Highest standards is not a reason to block the business while you polish.
It also shows up in the thinking. The best calls come from writing: a press release for the vision, a FAQ for the hard questions, a short narrative that tests the assertions with data. Inviting scrutiny on the writing is how the work gets better. Sloppy prose is sloppy thought.
Calibration
Under, just right, over.
Look for a defect they stopped from going downstream, and a launch they held for a real reason. "I have high standards" is a slogan. Perfection that never ships is the over move. Shipping junk and calling it Bias for Action is the under move.
| Situation | Under | Just Right | Over |
|---|---|---|---|
| A standing meeting that always runs long | Lets the meeting keep the same shape. People stop preparing because nothing about it ever changes. | Changes one part of the meeting this week, such as the timebox or the order, and watches the next few sessions before changing anything else. Protects the time the meeting is supposed to save. | Redesigns the meeting every few days. The team spends more energy on the ritual than on the work the meeting should unblock. |
| Handing work to the next person | Passes a messy draft and assumes they will clean it up. The same gaps appear again one step later. | Removes the defects they would not want to inherit, and names the one leftover risk. Leaves the next person real work, not a cleanup they did not plan for. | Polishes every edge before anyone else can see it. The handoff is late, and the next person has no time left to react. |
| A workaround the team has used for months | Treats the workaround as the real process. Never checks whether a cleaner path is now cheap enough to take. | Keeps the workaround through this week's deadline, then makes a small change that removes one painful step. Does not wait for a full rewrite. | Stops the week's delivery to replace the whole workaround. The original work slips, and the new path is only half built. |
| Reviewing a teammate's draft | Approves it to stay out of the way. Leaves known gaps for someone later to find. | Flags the few issues that would hurt the reader or the next step, and lets the rest go. Says what the bar is so the next draft starts higher. | Rewrites the draft in their own voice. The author spends the rest of the week on taste, not on the substance that still needs work. |
| A problem that was fixed and came back | Applies the same patch and moves on. Does not ask why the first fix did not hold. | Restores the immediate path, then spends a short block on the condition that let it return. Leaves a note so the next person does not repeat the shallow fix. | Opens a rewrite of the whole area. The original problem stays open longer, and other work waits on a redesign nobody needed this week. |
| Picking what to improve this week | Says the team is too busy to change anything. Next week looks exactly like this one. | Chooses one friction the team actually felt and makes a small, reversible change. Guards delivery time so the improvement does not become the project. | Starts several improvement threads at once. Meetings fill with process talk, and the customer-facing work slips. |
| Showing a new person the current process | Walks them through the current steps as if they are finished. Does not name the rough edges or invite a better way. | Teaches the path that works today and points out one place it still hurts. Asks them to try a small fix only after they can do the job. | Hands the new person a redesign of the process in their first week. They cannot yet do the work, and the old path has no owner. |
| The day after something ships | Closes the ticket and does not look at what was clumsy in the last mile. The same scramble will happen next time. | Writes down one step that was harder than it needed to be and does a small cleanup while the memory is fresh. Does not reopen the whole release. | Holds a long wash-up and starts a quality program. The next release waits while the team debates standards. |
| Handing work to the next team | Sends the work with known gaps and lets the next team find them. | Fixes the issues that would become someone else's defects, and names what is still open. Leaves polish that would delay the handoff more than it helps. | Holds the package until every edge case is clean. The next team waits and the shared date moves. |
| The same bug comes back | Applies last week's patch again and closes the ticket. | Fixes the symptom and changes the check or step that let it through, even if that costs another day this week. | Stops the team's other work to redesign around one returning bug. Commitments for the rest of the week slip. |
| Reviewing a teammate's draft | Skims and approves so the author can stay on the calendar. | Calls out the few places that would actually lower the bar, and says why. Leaves wording nits that do not change the outcome. | Rewrites the draft in their own voice. The author stops owning it and the review sits in the way of the week. |
| The team says the bar is too high | Drops the standard so the argument ends. | Keeps the parts a customer would notice, and cuts scope or sequence so the team can still hit that bar this week. | Treats every objection as a lack of will and refuses to cut anything. People work late and the work still ships late. |
| Closing a ticket | Marks it done while a known defect is still in it for the next person to find. | Fixes what would travel downstream, writes down the leftover risk, and only then closes it. | Leaves the ticket open for cosmetic leftovers no customer will see. The board fills with work that is already good enough. |
| A new person ships their first work | Lets them send it out without showing what done looks like on this team. | Walks through one recent example of the standard and reviews their first piece against it, while still letting them finish the week. | Makes them redo the first task until it matches a senior's taste in every detail. They lose the week and still cannot tell which parts matter. |
| A partner team's deliverable is rough | Takes it as-is to keep the relationship easy. | Sends it back with the specific gaps that would turn into defects here, and a way to land it this week. | Rejects the whole package and rebuilds it inside. The partner stops trying and this team now owns their work. |
| A process already works but creates rework | Leaves the clumsy step alone because it has not failed loudly this week. | Tightens the one step that keeps producing rework, and leaves the rest until that change sticks. | Redesigns the whole process in the same week. People spend their time learning new steps instead of doing the work. |
Examples
What it looks like in the work.
Stop the line
A defect is about to travel to the next team, or to a customer. The just-right move is to stop the line, fix the cause, and start again. Passing it downstream so your date looks clean is the under move. Look for a time they pulled that cord, and what stayed fixed afterward.
Hold the launch two days, not two months
A bad bug is a reason to slip a couple of days. A typo in an appendix is not. Highest standards without triage is fear. Bias for Action without standards is hope. Name the defect that was worth the hold, and the defects you shipped on purpose.
The writing is the thinking
A short narrative, a press release, and a FAQ will expose a weak idea faster than a slide deck. If the numbers in the appendix do not match the body, the thinking is not done. Look for how they invited other people to tear the document apart, and what got better because of it.
In the role
Individual and manager.
Individual
Your docs match themselves. You do not send a known defect to the next person. You can say which standard you raised after you hit the old one. You get annoyed by "good enough" when the customer will feel it, and you can still ship when the leftover work is truly leftover.
Manager
You set goals the team is proud to hit, not ones they sandbag or ones they cannot believe. You review work. You name the miss. You fix the system so the miss cannot recur. You do not confuse relentless standards with a culture where nobody will take a swing.
Go deeper
Questions that make the principle concrete.
- What defect did you stop from going downstream, and what did you change so it could not recur?
- When did you hold a launch, and which defect was worth the hold?
- Which standard did you raise after the team had already hit the old one?
- Where are you accepting a lower standard right now because the date is close?
- How do you decide a goal is a stretch instead of a sandbag or a fantasy?
- When the first answer on a problem was tidy, what did you do next?
- What did you refuse to compromise on, and who was the customer of that refusal?
- How do you review work without creating a team that is afraid to start?
- Where do the numbers in the document disagree with each other, and what does that tell you?
- How do you keep Bias for Action from turning into an excuse to ship junk?
From the blog
Writing that goes deeper.
- How Meeting/Not-Meeting Goals relates to Earn Trust and Insist on Highest Standards
The bar is real only if you set a goal and hold yourself to it in public.
- Details Matter in Presenting Narratives
Sloppy prose is sloppy thought. The writing is the thinking.
- Do Your Job – Don't Use Placeholder Text
Shipping lorem ipsum is sending a defect downstream on purpose.
Related