Why Companies Make Obviously Sh***y Decisions

Bad Process, Bad Processing, and Bad Incentives—and the Critical Differences Between Them

Image Source: ChatGPT.

A few days ago, an analyst friend sent me a text out of the blue: “why do companies do things that are obviously unlikely to work?” It persuaded me to dust off an essay I wrote years ago, but never published.

P.S. The *** in the title stands for “odd.” The word is “shoddy.” This is a family publication!

---

I was a Wall Street analyst for over a decade. A critical part of the job is to pass judgment on the decisions companies make. Truth be told, I always had my doubts that we were qualified to do so.

That’s because I had a very romantic conception of how big companies make decisions. Surely, they had clear objectives, good analysis, rigorous process, the right people in the room, and open debate. Plus, they had virtually unlimited resources, including access to vast amounts of non-public information and all sorts of outside counsel and consultants. Senior executives are, more than anything else, professional deciders. They must be really good at it. Right?

When I joined Time Warner, I was disabused of this romanticism early and often. To be clear, some of the smartest and most driven people I’ve ever met worked at Time Warner. Unlike many big companies, it also had a culture of open debate, fostered by our CEO Jeff Bewkes. We made a lot of good decisions. But we made a few bad ones too. And not bad in the sense that they ultimately failed to achieve their objectives, but decisions that seemed bad at the time.

Why? And so what?

Tl;dr:

• In my experience, almost all obviously bad corporate decisions occur for at least one of three (non-mutually exclusive) reasons: 1) bad process (errors of analysis); 2) bad information processing (errors of perception); and 3) bad incentives (errors of intent).

• Bad process is self-explanatory. It happens when companies move too fast or rely too much on intuition.

• Bad information processing occurs because companies are run by people, and people are flawed. Behavioral economics provides a useful framework for understanding these errors. Without getting bogged down in long lists of cognitive biases, the fundamental problem is that people are biologically primed to jump to conclusions and, once they have, not evaluate them objectively.

• Misaligned incentives, otherwise known as the principal-agent problem, occurs when the deciders put their interests first.

• Startups and founder-led companies are most prone to #1 and, as a result, #2. Aging companies with poor growth prospects run by professional managers are most vulnerable to #3.

• Understanding the underlying cause of bad decisions matters. #1 and #2 can be fixed. #3 is a sign of cultural and ideological rot and, often, of a company’s impending decline. And, short of a wholesale change in leadership, it usually can’t be fixed.

---

---

A Non-Trivial Question

If you pay attention to the decisions that companies make for long enough, you are bound to see some real head scratchers. That will never work! When I talk about bad decisions, I don’t mean decisions that don’t work out. In Thinking in Bets, Annie Duke coined the term resulting: judging the quality of decisions on the outcome, not the process. So, I don’t mean that. I mean decisions that are clearly bad when made.

One of the worst decisions I saw at Time Warner happened almost the day I showed up.

It was early 2008 and I had just joined the company. In the days prior to my first official day running Investor Relations, I started to get a lot of email traffic about something called “Project Birmingham.”

Project Birmingham was the codename for our acquisition of a company called Bebo, the third largest social network at the time, for $825 million. We announced it my first day in charge of IR. I spent the entire day on the phone trying to mollify furious institutional investors, who immediately thought it was a horrible idea. Plus, this was new-CEO Bewkes’ first major transaction and they interpreted it as an indication that he would be a bad allocator of capital. (He ended up being a great steward of capital: he did the unthinkable in media and got smaller, by separating Time Warner Cable, AOL, and Time Inc.; repurchased about one-third of the stock; and sold the company before the downturn of big media stocks. But it was a rough start.) The stock got pummeled, losing more than $2.5 billion of market value over two days, several multiples of the purchase price. One investor told me that the deal failed her “eight-second” rule—and if she knew in the first eight seconds that a deal was bad, then it would end up being really bad.

She was right. Two years later AOL, by then an independent public company, sold Bebo for what was reportedly less than $10 million.

Why do good (and even not-so-good) companies make obviously bad decisions? It’s an increasingly important question for a bunch of reasons.

Over the last few decades, the largest companies have gotten ever larger relative to the broader economy. Maybe this growing concentration is a natural economic phenomenon, maybe it’s due to lax antitrust enforcement, maybe it’s because of the winner-take-most dynamics of the internet. Whatever the reason, it’s demonstrably happening, whether based on assets and revenue (Figure 1) or market capitalization (Figure 2).

Figure 1. Rising Concentration of Assets and Revenue

Source: The University of Chicago Becker Friedman Institute.

Figure 2. Rising Concentration of Market Cap

Source: S&P, via Osborne Partners.

As the biggest have gotten bigger, naturally their influence has grown too. The largest companies create the largest externalities and will have an enormous effect on climate change, income inequality, and privacy and the development of life-altering technologies, like AI, robotics, and synthetic biology.

Like I said, how companies make decisions is important.

In the years after I left Time Warner/Warner Media, I spent a lot of time trying to reconcile my naïve preconceptions about corporate decision-making with the reality I saw both within Time Warner and at other companies. My conclusion is that most bad corporate decisions are due to at least one of three (non-mutually exclusive) reasons: bad process, bad information processing, and misaligned interests. The first describes errors of analysis; the second errors of perception; and the third errors of intent. The distinctions between them are important.

Bad Process

The first cause of bad decisions is bad process, which often means lack of process altogether.

One of my favorite scenes from the HBO show Succession is probably not high on other fans’ lists. It’s from the “Argestes” episode in Season 2, when the Roystar-Wayco gang is at an exclusive conference that is clearly modeled after the Allen & Co. Sun Valley Conference.

Tom Wambsgans, CEO of the fictional ATN news network, is preparing to roll out a new slogan and branding campaign at the conference: “We’re Listening.” Cousin Greg runs up to him shortly before his presentation and breathlessly tells Tom that “legal and comms” thinks he needs to kill the new tagline. He explains that Roystar-Wayco has embedded technology in its set tops that eavesdrops on customers’ conversations. “So, it turns out, we actually are listening.” Tom goes to his scheduled talk and makes up an entirely new campaign on the spot. It’s a recurring theme throughout the series: just how many important decisions are made up off-the-cuff.

Like anything else, too much process is counterproductive. In the real world, executives always have to make decisions with incomplete information. Sometimes, they need to move very fast. Overanalyzing the potential outcomes of important decisions will paralyze an organization. Overanalysis also lends itself to false negatives; since it is likely to unearth more reasons not to do things than do them, it probably causes missed opportunities. But having a clear and repeatable process for important decisions is crucial. That includes goals (what is the definition of success and how will we measure it?); governance (who decides?); shared fact patterns and analysis; a systematic way to incorporate diverse viewpoints; decision-making criteria; and some sort of postmortem to evaluate (and improve) the process itself.

Often, there’s none of that. There’s some superficial analysis, window dressing meant to support a decision that’s already been made. An executive has a hunch, has already made up his mind, and has no desire to do the work to arrive at an objective decision or encounter evidence that contradicts what he already decided. Dissenting voices either aren’t in the room or are afraid to speak up.

I worked with an executive who, when we first met, proudly told me “For me, it’s not about analysis. I go with my gut.” There’s an important role for gut and intuition in the decision-making process, but it shouldn’t be the only input.

Bad Information Processing

The second, related cause of bad decision-making is failures of information processing—i.e., failure to objectively evaluate the information at hand. The reason this happens is because companies are run by people, and people are flawed. The whole field of behavioral economics is dedicated to explaining these flaws. It provides a useful framework to understand these types of mistakes, with a caveat.

Behavioral economics is relatively new. It only emerged as a recognized academic field in the 1980s-90s and arguably entered mainstream thinking when Daniel Kahneman published Thinking, Fast and Slow in 2011. I was very excited when I first read it, because it offered a new way to think about decision-making. One of the primary assumptions underlying classical and neoclassical economics is that decision makers are rational. That’s handy for building mathematical models and constructing a rough framework for how markets work, but if you’ve ever met other people you also know that it’s wrong. People are quite often irrational. Behavioral economics explained this irrationality by defining a litany of cognitive biases and heuristics (mental shortcuts) that are now part of the business lexicon: confirmation bias, loss aversion, prospect theory, hindsight bias and more. All of these are essentially failures to objectively evaluate information and causes of bad decision-making.

Behavioral economics essentially describes failures to objectively evaluate information.

As a framework for understanding bad decisions, however, it has limitations. Type “behavioral economics” into Google and you’ll find long lists or schematics of cognitive biases, sometimes comprising dozens or hundreds of interrelated and overlapping concepts. A common (and I think legitimate) criticism of behavioral economics is that it’s just a “pile of quirks.”

For our purposes, rather than focus on specific biases, it’s more helpful to distill it down to the fundamental idea. So, here’s my one-sentence summary of behavioral economics: we are biologically primed to jump to conclusions and, after we reach those conclusions, not evaluate them objectively. To paraphrase Hillel the Elder, the rest is commentary.

Here’s behavioral economics in one sentence: we are biologically primed to jump to conclusions and, after we reach those conclusions, not evaluate them objectively. You see a lot of that in corporate decision-making.

Properly unpacking the evolutionary and biological root causes of cognitive biases is an essay in itself. Here’s my elevator version:

• We evolved to optimize cognitive efficiency, because thinking is metabolically expensive. That’s why we take mental shortcuts.

• We are biologically averse to ambiguity, because slow decision-making could be deadly in the savannah. That leads us to quickly jump to conclusions.

• We evolved to avoid threats for obvious reasons, leading to loss aversion and cognitive biases like prospect theory, which causes people to hate losing much more than they like winning.

• We are social animals, which leads to conformity, aversion to confrontation, acquiescence to authority, and other biases.

• We have strong self-identities, which is evolutionarily beneficial because it increases the likelihood we’ll act in our own interests. But our tendency towards self-affirmation also leads to many biases, like confirmation bias, hindsight bias, etc.

The point is that our biology often gets in the way of good decisions. You see a lot of behavioral economics at work in poor corporate decision-making. Bad deciders jump to conclusions. And, once made, teams seek out evidence that supports the decision and ignore evidence that challenges it.

Misaligned Interests

The most insidious of these causes is misaligned interests, or what is also called the principal-agent problem. The interests of the “agents”—senior managers and the board—are misaligned with those of the “principals,” the various stakeholders who care about the long-term health of the company.

The most insidious cause of bad decisions is misaligned interests.

For the first two causes of poor decisions described above, errors of analysis and errors of perception, the decider may want to make the right decision for the company, but something is standing in her way. In this third case, the decider puts her own interests first.

Sometimes, the conflict of interests is obvious. The most pervasive principal-agent problem—so obvious that people don’t even think about it—is short-termism, driven by a desire to appease Wall Street. Senior managers are often evaluated by the performance of their stock, which is in turn correlated with their ability to deliver quarterly earnings that at least meet or, hopefully, surpass Wall Street expectations. Most executive compensation plans are also highly dependent on current year financial performance. As a result, senior managers often do things to prioritize near-term results at the expense of the long-term health and competitive position of the business.

Misaligned interests can be obvious and overt or implicit and subtle.

There are many other overt examples. Maybe some otherwise worthless financial engineering would boost a key financial metric (like free cash flow!) that’s an important part of the named officers’ compensation. Maybe a CEO wants to acquire another company of dubious strategic value because getting bigger would change the comparables used by compensation consultants and justify a higher paycheck. Maybe a CEO shies away from a decision that has a good chance of paying off long term but would pressure near-term earnings—because he’s worried that an activist investor or hostile bidder could show up and make his life difficult or even take control of the company and fire him and his friends. Maybe he wants to buy something so that he can recast the financial statements to distract from the poor performance of the core business. Maybe he wants to launch a new business venture solely because of optics, to persuade the press or board that he is doing something. Maybe she is retiring in two years but the tough decisions today wouldn’t pay off for three or five, so there’s no upside to making life more difficult now.

Other times, the conflicts can be implicit and quite subtle. Maybe the senior executive just doesn’t have the stomach for risk or change. Maybe a decision to, say, reorganize the company would necessitate hard conversations with longtime lieutenants. Maybe the strategic decision at question doesn’t align with the CEO’s identity (“I am an acquirer of assets, not a seller.”)

They are More Common in Some Types of Organizations than Others

These three classes of errors are more common in some types of organizations than others.

Startups and early stage companies are much more likely to fall prey to #1 and, as a result, #2. They have to move quickly and, as young companies, there’s a good chance they don’t have great decision-making processes in place.

Founder-led companies, whether early stage or large public companies, are prone to #1 for another reason. They are often cults of personality, in which many are reluctant to challenge the founder. As a group, founders are generally highly confident people. The more successful they get, the more this self-perception goes up and the more their humility evaporates. But they are also far less likely to fall prey to #3. That’s because founders often entwine their personal identities with the success of the organization. Steve Jobs was Apple. Elon Musk is Tesla. Jeff Bezos is Amazon. The principal-agent dichotomy isn’t a conflict because there’s no dichotomy. In The Founder’s Mentality, Chris Zook and James Allen define that eponymous mentality as having a clear mission, an owner mindset, and a relentless focus on the customer. Owners don’t think like agents.

Mature companies run by professional managers are where #3 rears its ugly head. I once worked with a senior executive who equated his business unit to a train. “It comes into the station, I get on, I ride for a few stops, I get off, and it keeps going.” There’s something healthy in that attitude, because it keeps the executive humble, but it also raises the risk that he also sees his own interests as divergent from the organization’s. Aging companies with poor growth prospects are the most vulnerable to managers who put their interests first.

Understanding the Causes of Bad Decisions is Important

So, here’s my answer to my friend’s question: why do companies do things that are very unlikely to work? Bad process, bad information processing, and/or bad incentives.

Seeing a lot of them is a red flag, but the reasons matter.

#1 and #2 are easier to fix, and in related ways. Good process can go a long way toward preventing cognitive biases from tainting decisions. There are several techniques, for instance, engineered specifically to override confirmation bias. Kahneman wrote about the importance of incorporating an “outside view” and accounting for base rates when evaluating decisions.¹ Cognitive psychologist Gary Klein devised the premortem, an exercise to project out to some point in the future, imagine that the initiative in question has failed, and work backwards to identify the reasons why.

A technique I have found personally useful for short-circuiting cognitive biases: asking what would have to be true for this option to work?

One of my favorite approaches comes from Roger Martin, which he laid out in Playing to Win. It consists simply of taking each option on the table and then asking, for each, what would have to be true for this to be the best choice? It’s a managerial version of the scientific method: the null hypothesis is that the strategic option will succeed and you try to disprove it by identifying the most important assumptions and testing them. Like the scientific method, the philosophical foundation of this approach is falsifiability—a hypothesis remains viable until proven wrong. This is very different from most strategic processes, in which the team comes up with some option and then sets out to prove why it’s right. This often causes the group to focus narrowly on whether the key explicit assumptions underlying the choice are accurate and overlook the hidden, implicit (and often, more important) ones. They also tend to seek out confirming information. By contrast, when the team tries to prove it to be false, they must make the implicit assumptions explicit and actively seek out information that might scuttle the plan.

But my goal here isn’t to provide an exhaustive approach to instituting good decision hygiene and overcoming cognitive biases. The point is that these first two problems can be fixed. The third one is much tougher.

As I mentioned above, misaligned incentives are not errors of analysis or perception, they are errors of intent. When they run rampant, the logic of the organization has been replaced by the logic of self-enrichment and self-preservation. It’s horrible for morale, as junior employees become disillusioned and demotivated. It’s usually an indication of cultural and ideological rot and a good sign that the company is in decline. It also usually suggests the board is asleep at the switch or in some self-dealing autocrat’s pocket. The only way to fix this problem is a wholesale change in leadership.

So, the next time you see a pattern of obviously dubious decisions, try to suss out why. Is the company just moving fast? Is it founder led and following the whims and intuitions of its leader (which have likely been pretty successful in the past)? Or are the senior managers just looking out for themselves? You can fix a process. But you can’t overcome a leadership team that is putting its interests over the company’s.

---

1. Base rates are the typical outcome for similar events. The base rate for restaurant failures, for example, is that about half make it to five years.