Why companies decline: the choices that turn a slip into a fall

The 3,400 salespeople Circuit City let go

On March 28, 2007, Circuit City let go about 3,400 of its store employees. The company called it "wage management": anyone earning well above the market rate for their role was out.

Anyone let go could reapply ten weeks later, at a lower salary. On paper it was a saving on payroll. On the sales floor it was the loss of the people who knew the products best.

Within weeks Circuit City was expecting a first-quarter loss, and a Deutsche Bank analyst wrote that the labor change had hurt service levels and was handing market share to Best Buy. In November 2008 Circuit City filed for bankruptcy.

When it filed, Circuit City blamed its vendors, its liquidity and the global economic crisis, not its payroll. Both stories can be true at once, and that is what makes this decision so hard to see from inside. A fall rarely has one cause. But it often has one choice that, looking back, took away something the company could not get back. Here, it was its most experienced salespeople.

What the fall is

The decline itself is often not a choice. Customers move, technology changes, a rival arrives with something better. What the company does next is a choice, and it is usually made while the old business still looks healthy enough to protect.

Why it is so hard

From inside, none of these choices looks like the one that ends a company. Three things make them hard to see.

The ways companies did it

The decisions on record take four shapes. Each protects the present at the expense of what comes next.

What 49 companies did

Here is every decision on record, placed by the year it was made and marked by how it turned out. Hover a mark, or open the table, for the company and the outcome.

Of the 48 decisions that can be judged, 34 were never undone, 13 were walked back, and 1 held only in part.

On this page "never undone" is not a compliment. It covers the choices that ended Circuit City, Pan Am and Silicon Valley Bank. They were never undone because by the time anyone wanted to, there was nothing left to undo them with.

Nokia's switch to Windows Phone is on record twice, told from two angles, so it appears twice on the timeline.

The walk-backs

Put the walk-backs side by side and they fall into three groups: products that flopped and were pulled, purchases that were sold or written off, and policies the company itself took back once the damage was done. One more choice was cut short by a court rather than by the company.

The flops were the cheapest to undo. Amazon cut the Fire Phone's contract price from $199 to $0.99 six weeks after launch and took a $170 million charge. Nintendo took the Wii U's idea, gave it a name people understood, and sold 155.92 million of them as the Switch.

The decisions that ended companies are mostly on the other list, marked never undone: Circuit City's layoffs, Pan Am's sale of its Pacific routes, Silicon Valley Bank's bond sale. A product can be pulled from the shelf. Staff you let go and routes you sell do not come back when you change your mind. Even a reversal can come too late: Borders took its website back from Amazon in May 2008 and still liquidated three years later.

A flop can be taken back. A fall usually can't.

What the share price did

For some of the decisions there is a clean market record: the company's total return over the three years after the decision, against the S&P 500.

Most of the marks sit below the line, which should surprise no one: these are companies in trouble. The useful lesson is in the exceptions to the story. Enphase's switch to solar-plus-storage in 2023 was never undone, and its revenue climbed back toward $1.47 billion in 2025, yet its shares trailed the index by 148.9 points over three years. Surviving is not the same as recovering.

NIO, the mark far above the line, is a company that came close to failing. It survived 2020 on a government-led rescue of RMB 7 billion, and recorded its first profitable quarter at the end of 2025.

How to run this decision

If the business that made you dominant is slipping, here is what the record suggests doing, in order.

1. Before you cut a cost, find out what customers get from it

   Payroll, service, a website, a studio: each is a line in a budget and also a reason someone buys from you. Ask what a customer loses if it goes, not only what the company saves. Circuit City saved on wages and lost its most experienced salespeople, and its customers went to rivals with better-trained staff.

   Seen in: Circuit City, 2007 · Google Stadia, 2021
2. Price the way back before you sell or hand over a core asset

   A route network, a sales channel or a customer relationship sold for cash is rarely there to buy back later. Pan Am sold its Pacific Division to a direct competitor for $750 million. Borders handed its online customers to Amazon and stayed on Amazon's platform for seven years.

   Seen in: Pan Am, 1985 · Borders, 2001
3. Make any acquisition answer the question the core business is asking

   If the core is shrinking, a purchase next door does not fix it. It adds a second business to run while the first one keeps shrinking. GameStop bought phone stores to escape its disc business, took $344.2 million of impairments on Spring Mobile, and then sold it.

   Seen in: GameStop, 2013 · AT&T, 2016 · Microsoft, 2013
4. Test with paying customers before you launch at full size

   A full-scale launch spends the expensive half of the bet to learn what a small test would have told you. Quibi built a full service around a thesis it had never tested on a paying customer. Crystal Pepsi was rushed out even though a food scientist had flagged a risk before launch.

   Seen in: Quibi, 2020 · Crystal Pepsi, 1992 · Amazon Fire Phone, 2014
5. Report the decline you have

   Pulling revenue forward buys a quarter and costs a reputation. When it comes out, and in these cases it did, the company must restate the past as well as explain the present. Xerox restated four years of results. Under Armour paid a $9 million penalty over $408 million of orders it had shipped early.

   Seen in: Xerox, 1997 · Under Armour, 2015
6. Write down what would prove the new rival is serious

   In 2007 Jim Balsillie of BlackBerry called the iPhone "kind of one more entrant into an already very busy space". In 2011 Nokia's chief executive wrote that the first iPhone had shipped in 2007 and Nokia still had nothing close. Decide in advance which number, by which date, would mean the threat is real, and what you would do then.

   Seen in: BlackBerry, 2007 · Nokia, 2011

Warning signs

What it has looked like when this decision was going wrong.

• The saving comes from the people or the service customers deal with directly.Circuit City, 2007
• A core channel is being handed to the rival that threatens it.Borders, 2001
• The new competitor is described as just one more entrant.BlackBerry, 2007
• Orders are being shipped early to protect a growth streak.Under Armour, 2015 · Xerox, 1997
• The launch is full size, and the idea has never been tested on a paying customer.Quibi, 2020

All 49 cases

Each case is one company at one decision: what was on the table, what it chose, why, and what followed, with every fact sourced.

Questions

Why do successful companies decline?

Rarely from one blow. In the cases on record, decline usually runs through a choice made while the business was slipping: cutting something customers valued, buying a business to escape the core, launching before the product was ready, or making the numbers look better than the business. At the time, the choice usually looked sensible.

What are examples of companies that declined after a strategic decision?

Circuit City let go about 3,400 of its best-paid store employees in 2007 and filed for bankruptcy in November 2008. Borders handed its website to Amazon in 2001 and liquidated in 2011. Pan Am sold its Pacific Division in 1985 and stopped flying in 1991. AT&T bought Time Warner in 2016 and handed it to Discovery in 2022.

Can a declining company recover?

Some do, and the turnaround guide covers how. In this record, many of the decisions that were walked back were failed launches, and for most of those the cost was a product, not the company. The decisions that ended companies were the ones that could not be walked back.

What are the warning signs of corporate decline?

Savings that come from the people customers deal with, a core channel handed to a rival, a new competitor dismissed as one more entrant, orders shipped early to protect a growth streak, and full-size launches of ideas no paying customer has tried.