There’s a pattern in corporate failure that’s hard to unsee once you notice it.
The engineers saw it coming. The branch staff flagged it. The middle managers knew. And somewhere between them and the people making the call, the signal got softened, delayed, or lost.
The answers existed all along. They just never reached the people who needed them.
Here are twelve cases — a very, very small sampling — and the four filters that keep bad news from traveling up.
Filter 1: Saying it plainly is career-limiting
The most common filter is the simplest. Delivering bad news upward is risky, so it gets sanded down on the way.
- Nokia (2007–2013). Middle managers knew Symbian couldn’t answer the iPhone. Saying so was career-limiting, so it was softened on the way up. Global smartphone share went from 50% to 3%.
- Target Canada (2011–2015). Staff knew the inventory data behind the launch was wrong. Reporting it red was unwelcome, so the dashboards stayed green and 133 stores opened. Two years later the business was gone: $2B+ written off and 17,600 jobs.
- General Electric (2016–2018). A culture of “success theater” meant bad news stopped traveling up. The board learned how bad GE Power was only after the CEO left — $193B in market value, and out of the Dow after 110 years.
- Wells Fargo (2011–2016). Branch staff flagged the sales quotas for years. It stayed inside the division; the board learned the scale of it from the news. $3B in penalties, 5,300 staff fired, and the CEO gone.
None of these organizations lacked smart people. They lacked a channel where the truth was safe to say.
Filter 2: The signal gets lost in summarization
Even when people speak up, every layer of the org chart summarizes — and every summary drops detail.
- NASA (1986). Engineers argued for hours against launching Challenger in the cold. Their objection was overruled and never reached NASA’s senior managers. Seven lives were lost, and the Rogers Commission found the launch wouldn’t have happened if they’d been told.
- Boeing (2014–2020). Late in development, MCAS was given four times its original authority, off a single sensor. The change never traveled — not to the manuals, not to the pilots, barely inside Boeing. 346 lives lost, $20B+ in costs, and a 20-month grounding.
- Equifax (2017). The patch notice went to a list that didn’t reach the team running the exposed system. Nobody above them knew it was still open. 147 million people affected; $1.4B to remediate and $700M in settlements.
By the time information reaches the top, it’s often late, partial, or gone.
Filter 3: Everyone agrees, nobody owns it
Sometimes the problem is known widely — and owned by no one.
- General Motors (2004–2014). Engineers documented the ignition defect for a decade. The internal report blamed the “GM nod”: everyone agrees, nobody owns it. 124 lives lost and a $900M criminal fine.
Knowing isn’t the same as acting. A signal that isn’t attached to an owner and a deadline is just a note.
Filter 4: Leadership is too far from the action
The last filter is distance — physical, organizational, or simply the distance of believing you already know the answer.
- BlackBerry (2007–2013). Engineers saw the app ecosystem coming. Leadership bet enterprises would never give up the keyboard. Global share went from 20% to under 1%.
- Kodak (1975–2012). Its own engineer built the first digital camera. Management shelved it to protect film. Kodak went bankrupt in the market it invented.
- Xerox PARC (1973–1981). Its own lab built the personal computer — screen, windows, mouse, network. Head office was 2,500 miles away and never grasped what it was holding.
- Circuit City (2007–2009). 3,400 of the highest-paid salespeople were cut to save wage costs. To store managers they were the people who closed the sales; upstairs they were a payroll line. Liquidated within two years, taking 34,000 jobs.
Why this gets worse as you grow
None of this requires bad people. It’s structural.
At 50 employees, a CEO hears most things directly. Somewhere past 150, that stops being true. Skip-levels become impossible to keep up, and every new layer becomes another filter:
- People are too busy to report upward.
- Sharing bad news is career-limiting.
- Important details are lost in summarization.
- It’s hard to separate signal from noise.
The cost shows up in three places: wasted time wrangling information instead of driving progress, late decisions made on partial facts, and costly mistakes and missed opportunities that were visible on your own frontline months before they reached you.
What the leaders who get this right do differently
One of the leaders I most admire, Kat Cole, has a principle I think about constantly: stay close to the people who are close to the action. The people on the frontline usually know what to do long before senior leaders do, because they’re closer to the customers and the operations serving them.
That isn’t soft, cultural stuff. It’s strategic. In practice, the leaders who avoid these failures tend to do five things:
- Ask, regularly — not once a year. Make listening part of the operating rhythm, on the questions that matter to the business: risks, blockers, opportunities, not just engagement.
- Make it safe to be candid. Real anonymity — enforced, not promised — changes what people are willing to say.
- Dig for specifics. “Things are slow” isn’t a finding. What’s an example? What’s causing it? What’s it costing us? is.
- Count, don’t anecdote. One person raising something is an anecdote. Twenty people raising it independently is a finding. You need to know which you’re looking at.
- Close the loop. Attach every real signal to an owner, a goal, and a deadline — and tell people what you did with what they said.
The lesson
Listen to the people closest to the action. It sounds obvious, and it is. The hard part is doing it at scale, continuously, in a way people trust — and making sure what you hear actually changes what you do.
That’s exactly what we built Elevea to help leadership teams do.
