← Field Notes
Operating doctrine4 min read27 responses

What are you still tolerating?

Most systems do not fail in one cinematic moment. They become expensive through a thousand accepted frictions.

On fire92
ACCEPTED FRICTION / 001

# What are you still tolerating?

Most systems do not fail in one cinematic moment. They become expensive through a thousand accepted frictions.

There is a peculiar kind of problem that never becomes urgent enough to deserve a war room and never becomes small enough to disappear. It lives in the organization like a bad knee. Everybody knows where it hurts. Everybody has learned how to compensate. Nobody expects it to heal.

The report is rebuilt by hand every Friday. The customer has to tell the same story three times. A project manager keeps the real schedule in a private spreadsheet because the official tool cannot be trusted. The person who understands the process cannot take a vacation without carrying the business in a backpack.

These are not inconveniences. They are tolerated design decisions.

Tolerance has a carrying cost

People are remarkably adaptable. That is one of our best qualities and one of the reasons bad systems survive. A capable person can create a workaround before breakfast. A generous person will quietly absorb a broken handoff rather than embarrass a colleague. A committed team will use memory, extra meetings, text messages, color-coded notes, and heroic effort to keep an operation upright.

The organization sees the result and concludes that the process works. It does not. The people work. The process is borrowing their intelligence, attention, and goodwill without recording the debt.

The carrying cost appears in places accounting rarely labels correctly: slower decisions, uneven service, exhaustion, errors caught late, customers who stop asking, talented people who become cynical, and managers who spend their best hours chasing information that should already be visible.

Make a tolerance inventory

Before rebuilding anything, inventory what people have normalized. Do not begin with software. Begin with sentences.

  • “Only Maria knows how to do that.”
  • “The status in the system is not the real status.”
  • “You have to call them after you submit it.”
  • “We keep a second list so nothing gets lost.”
  • “It is faster if I do it myself.”

Each sentence is a flare. It points toward a recurring tax on the operation. Some taxes are worth paying. Reality contains ambiguity, judgment, and exceptions. The goal is not to automate every human moment. The goal is to stop spending human judgment on problems that no longer deserve it.

Ask what the workaround protects

Every durable workaround protects something: speed, a customer relationship, a deadline, a manager from bad news, or an employee from a tool that punishes honesty. Removing the workaround without understanding its purpose is how process improvement makes the work worse.

Suppose a coordinator keeps a private tracker. The lazy response is to demand everything move into the official system. The useful response is to ask why the private tracker exists. Perhaps it shows dependencies the official system hides. Perhaps updating the official record creates notifications nobody wants. Perhaps its statuses were written for executives rather than for the people moving the work.

The private tracker is not insubordination. It is user research.

Convert irritation into a specification

A complaint is only the beginning. “This is annoying” does not tell us what to build. Convert the irritation into an operating requirement.

  1. Name the recurring moment.
  2. Measure the compensation required to survive it.
  3. Identify the outcome the workaround protects.
  4. Define the honest signal the system should reveal sooner.
  5. Give someone a clear next move when that signal appears.

This changes “we need a better dashboard” into “the owner of a delayed inspection must see the blocking document, responsible person, last action, and next escalation without opening five systems.” That is a system a team can design and test.

Choose one expensive tolerance

Large transformations often fail because they begin by declaring everything broken. Choose one tolerance that is frequent, visible, and expensive. Fix it completely enough that the organization feels the difference. A strong first win does more than save time. It changes what people believe can be improved.

Evidence matters. If the change reduces cycle time, missed handoffs, duplicate entry, customer repetition, or emergency escalations, make that visible. Improvement without proof becomes folklore and is eventually reversed by someone who never experienced the original pain.

The question is not whether your people can endure the system. They already have. The question is what becomes possible when they no longer have to.

Ask the question every week

Ask one question in a recurring team meeting: What are we still tolerating?

Do not turn the answer into an immediate assignment for the person who raised it. That teaches silence. Record the friction. Look for repetition. Assign ownership deliberately. Close the loop visibly.

Some answers will be petty. Some will be impossible. Some will reveal that the organization has been spending thousands of dollars and hundreds of human hours protecting a flaw nobody actually wants.

That is the opportunity.

Member discussion

What does this change for you?

The article is the opening argument. The useful room begins when someone brings evidence, disagreement, or an application from the real world.

READING IS PUBLIC · THE ROOM IS FOR MEMBERS

The conversation starts with identity.

The article is public. The room behind it is not. Members see every response, add to the discussion, and carry a durable identity across the work.

Enter the room ↗
DO NOT STOP AT ONE IDEA

Continue through the system.

The next article is selected because it extends, challenges, grounds, or applies the argument you just read.