Part of the Pressure Intelligence™ framework
Pressure Pattern #13

Check Engine

On the early signals that make a course correction cheap, and the threshold, set when a decision is made, that says when to stop and look again.

October 27, 2026
Check Engine

Pressure: A warning is easy to explain away when its meaning was not agreed in advance.

Plenty of people have driven a car for months with the check engine light on. The light has two settings, on and off. The first time you see that little light, you might feel a knot in your stomach and wonder what’s going on, and put it on the mental to-do list with high urgency. But a month or two later, if the light is still on, the car still starts, and you are still getting to work just fine, that light can easily fade in importance and just become part of the dashboard as the original urgency you felt begins to feel normal.

Sometimes the check engine light is reporting something small. Sometimes it is the first sign of a problem that, left alone, can do enough damage to ruin the engine.

The longer a check engine light stays on, the more expensive the answer tends to be.

The light works at all because a product team had decided, long before you bought the car, what conditions should turn it on. Organizations seldom do that for their own decisions. I learned that the hard way, in a stretch when the business seemed to be taking off.

We had just closed a large investment. Competitors were moving, the market felt like it was opening, and the executive team (myself included) and the board decided to hire ahead of revenue. It felt like the moment to strike and the window of opportunity might not be open for long.

I was still uneasy. Spending money we could not yet see didn’t sit well with me. But our collective response was to have more confidence that the revenue would come. Success depended on a few large outcomes, and each of them looked a little closer than the month before.

What I did not do was ask the more useful question early: what conditions would have to be true, and by when, for us to revisit the strategy?

The question of pace and how much runway we had came back each quarter, and each quarter it was settled the same way, by feel. One quarter without meeting the revenue target was a warning light, and each successive one arrived along with a good reason to keep going: one of the big outcomes we were after looked closer than it had before, another was starting to move, and the market still looked like it was opening. With no line drawn ahead of time, those reasons had nothing to be measured against. The outcomes stayed close, quarter after quarter, yet frustratingly, they did not land.

Eventually, we hit a wall and were forced to make a drastic change which was significantly more painful than a mid-course correction would have been. I only realized after months of reflection that the tripwire had not been set at the beginning. We didn’t agree as a team which conditions would have to be true in order for us to change course and revisit the original decision.

Pattern: Warnings that arrive one at a time tend to get answered one at a time.

Identify the Pattern

Early signals in an organization share three features that make them easy to drive past.

  • They are small. A quarter that comes in short, a milestone that slips two weeks, a customer who asks the same setup question as the last one. Taken alone, each sits inside the normal wobble of a plan.
  • They come with an explanation. The deal moved to next quarter. The partner was late. The person raising a signal usually brings its cause along, and the cause is usually real. A true explanation makes a signal feel resolved.
  • They arrive separately. Different people, different meetings, different forms. The pattern shows up in the sum, and the sum tends to land outside any one person’s line of sight.

A fourth feature turns the first three from a nuisance into a cost: few signals come with a threshold agreed ahead of time. When nothing was agreed in advance about what a signal would mean, each new data point gets judged on its own, by people who are already invested in the plan. Seen alone, it looks survivable.

Karl Weick and Kathleen Sutcliffe studied organizations that have to operate reliably in conditions where small errors can grow quickly, such as nuclear power plants and the flight decks of aircraft carriers.[1] The first important habit they describe is a preoccupation with failure: these organizations treat a small lapse as a possible symptom of something wrong in the wider system, and they go looking for what it connects to. That habit runs against ordinary management instinct, which treats a small lapse with a good explanation as a closed item. In other words, when the light comes on, look for what it might connect to before deciding it is nothing.

The other half of the pattern is about people. A warning reaches a leader when someone chooses to bring it forward, and Mark Keil and colleagues set up an experiment to study that choice.[2] Participants in a role-play learned that a software project was in trouble and decided whether to report it. When the workplace described in the scenario was more open to bad news, people felt more responsible for reporting it, and the more responsible they felt, the more willing they were to speak up. The participants were graduate business students acting out a scenario, so treat the result as a tendency rather than a measurement of real workplaces. The problem the study starts from will sound familiar to anyone who has run a large program: when a big project goes wrong, it can take weeks or months for senior leaders to hear the full story.

Put the two together. Small signals get explained away where decisions are made, and the way they are received teaches the people carrying them whether the next one is worth escalating.

When a small warning gets a “noted” and nothing more, the next one is a little less likely to be raised.

What It Costs

The decision window. Early, the options are wide and cheap: slow a hiring plan by a quarter, try a different approach, move a date by a week. Late, they narrow to the expensive ones, and each of those lands on people, customers or commitments.

The accuracy of the plan. A plan that absorbs explained shortfalls one at a time keeps reporting green for longer than the work underneath it deserves. The people steering by it are steering by an older picture than they realize.

The quality of the argument. Without a threshold, a disagreement about pace or risk does not get settled. It gets reargued each time a new number comes in, and the side with more momentum tends to win, whatever the number says.

Reframe the Mindset

The assumption that gets in the way is that small problems are noise, and that a leader’s job is to keep the team on the plan instead of chasing each light that comes on. There is something to it. A team that stops for each flicker gets very little done, and a company that brakes at the first soft quarter can miss the window it raised money to reach.

A more useful understanding is that a small warning is the problem at its lowest price. The skill lies in telling which lights connect to the engine, and that is far easier when the meaning of each light was agreed before anyone needed it to say something comfortable.

A small warning costs a conversation. Found late, the same problem costs missed targets and lost opportunities.

Transform the Tension

Here are three things to try. Each one changes what a signal means, or what happens when one arrives, while the work keeps going.

Set the tripwire while the decision is fresh. At the moment you commit to a direction, write down what would make you reopen it: a number, a date, and what you would do if the number has not arrived by the date. It takes a few minutes, when the team is still clear-eyed about the risk. It turns a later argument about feel into a check against something the team already agreed, and it gives the person who was uneasy a way to be constructive instead of being viewed as overly cautious.

Read the warnings side by side. Once a month, put the small items from separate reviews on one page: the soft quarters, the slips, the workarounds, the questions customers keep asking. Then ask what they have in common. Three quarters of “almost” look different on one page than they did one review at a time.

Answer the first report with a question. The response to the first small warning sets the tone for the rest. Before offering a fix or a reassurance, ask: “If this turned out to be bigger than it looks, what else would we expect to see?” The question treats the person as a source of information, and it puts them to work on the connection you are trying to find.

Shift: Which light on your dashboard has been on long enough to look normal?

Reflection (Questions to Ponder)

  • Which of my current commitments has no condition written down that would make me reopen it?
  • What small issue have I heard about more than once, from different people, in the last month?
  • When someone last brought me an early warning, what did my response teach them about bringing the next one?
  • Where have I been uneasy about a decision without turning the unease into a question the team could answer?

Connection (Questions to Discuss)

  • What would have to be true, and by when, for us to change course on our biggest bet?
  • What are we seeing in more than one place that might be the same problem?
  • What would you have raised sooner if it had been easier to raise?

Subtraction (Things to Stop)

  • Stop committing to a direction without naming what would make you reopen it.
  • Stop closing a warning because its explanation is true. A true cause can still be a symptom of something larger.
  • Stop asking for certainty before someone raises a concern. What they have seen is enough to start with.
  • Stop answering an early warning with reassurance alone. “We’re fine” ends the conversation the warning was trying to start.

Action (Things to Try)

  • Pick your largest commitment this quarter and write one tripwire for it: the signal, the date, and the response. Share it with the people who will be watching for it.
  • List the shortfalls, slips, workarounds and repeated customer questions from the last eight weeks on a single page. Circle the ones that might share a cause.
  • Pick one recurring meeting and add a standing item: what small thing has come up more than once?
  • Ask the person closest to the work what they are watching that has not reached you yet.

What would make you reopen your biggest decision, and have you written it down?

Footnotes

  1. Karl E. Weick and Kathleen M. Sutcliffe, Managing the Unexpected: Sustained Performance in a Complex World, 3rd ed. (Hoboken, NJ: Wiley, 2015), chapter 3, “Principle 1: Preoccupation with Failure.” The book draws on research into high-reliability organizations; the application to business decisions here is mine. ↩︎
  2. Mark Keil, H. Jeff Smith, Suzanne Pawlowski, and Leigh Jin, “‘Why Didn’t Somebody Tell Me?’: Climate, Information Asymmetry, and Bad News about Troubled Projects,” The DATA BASE for Advances in Information Systems 35, no. 2 (2004): 65-84. A role-play experiment with 122 graduate business students, manipulating organizational climate and information asymmetry; it describes behavior in a scenario rather than in observed workplaces. ↩︎

About the author

Mark D. Benson
Mark D. Benson
Founder of Resonant Vector™
Resonant Vector
Full profile →

Mark D. Benson is the founder and principal writer of Resonant Vector, the home of Pressure Intelligence, and the host of Pressure Rising™. He is a product and systems leader who has spent his career running global teams, complex platforms, and consequential decisions under real constraint.

His career sits where strategy, technology and large human systems meet, in places where high stakes are the norm.

Resonant Vector is his commitment to learning in public and sharing what helps leaders and teams perform under pressure.

More about Mark can be found at markdbenson.com.

Subscribe to the newsletter

Clear thinking for leaders and teams navigating constant pressure, complexity, and change.

    We respect your privacy. Unsubscribe at any time.