Transcript

(Editor's note: transcripts don't do talks justice. This transcript is useful for searching and reference, but we recommend watching the video rather than reading the transcript alone! For a reader of typical speed, reading this will take 15% less time than watching the video, but you'll miss out on body language and the speaker's slides!)

[CLAPPING] I made my first proper decision at the age of 25. In the early 1990s my younger sister went to Russia for a couple weeks with a group of teenagers. And my dad, who was visiting us in Norway at the time, went along as a chaperone.

And the kids came back safe and sound. My dad wasn't with them when they returned. He had met a woman, fallen in love, and decided to stay in Russia.

I saw him a couple years later when he and my Russian stepmother came through Norway on their way to the US. This was right around the time when I accidentally moved to France for seven years.

[LAUGHTER]

In that time, among other things, I worked as a waitress. I was a receptionist at an investment bank, an extra on some movie sets, an archivist for the accounting department of a very swank nightclub. I was part of a silent acting troupe in the gardens of the Palace of Versailles, a typist at a legal firm, and a bilingual secretary for insurance adjusters who were specializing in jewelry store heists.

I also tried out for one of the top contemporary circus schools in Europe four years in a row. I failed every time. This is why I'm here to talk to you about decision making, because when I made my first decision at the age of 25, it was life changing. Not because the quality of the decision was so high, but because the bar was so low.

[LAUGHTER]

This decision is immaterial. I'm not going to tell you what it was.

[LAUGHTER]

When I first started working at GitHub a couple years ago, I stumbled across an unmonitored support inbox where event organizers were asking for money. And it wasn't really anybody's fault that these people weren't getting a response. The inbox had gotten lost in some organizational shuffle.

No matter how unintentional this was, it was just a little bit heartbreaking that we had managed to take something that was inherently full of warm fuzzies and turned it into a source of frustration and disappointment.

Now, it's not that we weren't spending our sponsorship budget. We were. But only on events that were explicitly brought to our attention in some way. So we had two very real problems. On the one hand, we had an enormous volume of incoming requests that were going unanswered. And on the other hand, we had serious potential for spending a significant chunk of money in ways that were severely biased. But if we wanted to reduce the amount of bias in these decisions, we had to evaluate all of the requests. And we had to make a call for every single one, yes or no. We had no guidelines for how to make this decision.

And I tried to kick the problem upstairs to where people had the authority to make decisions about money. And they bounced it right back down to me in one of those classic, don't bring me problems, bring me solutions kind of moves. And they asked me to produce an event sponsorship strategy.

So I started doing research. All the history of sponsorships at GitHub. I tried to talk to anyone who had ever been involved or knew anything about it, trying to come up with the parameters for this decision. But so long as an event wasn't actively disqualified, I couldn't find any basis for selecting one community event and rejecting another. We had a mountain of events requests. And we had no way to decide.

To be honest, I was maybe not the best person to be trying to solve this problem. Fundamentally, the way that I thought you made decisions was that you take whatever is in front of you, whatever presents itself, and you do a gut check. But this is not deciding. This can hardly even be characterized as choosing. Choosing assumes that you're choosing between two things. And choosing between yes and no is not even that. It's choosing between one thing, the thing in front of you.

The scariest part of this is that it's not just me. Right. This is a really common anti-pattern. There's research that suggests that this is extremely common even among well-educated adults who are literally paid obscene amounts of money to make decisions.

So I'm going to give you an example. Quaker Oats made two high profile acquisitions about a decade apart. In the 1980s they acquired Gatorade for $220 million. And in the 1990s they acquired Snapple for $1.8 billion dollars.

Now, the most striking similarity between these two acquisitions is not that they're both beverages. It's the decision making process that led to the acquisitions. In both cases the CEO tasted something, and he liked it, and the board agreed to buy it.

[LAUGHTER]

The most striking difference between these two acquisitions is the outcome. The Gatorade acquisition was a spectacular success. Within a short amount of time the brand was valued at an estimated $3 billion. The Snapple acquisition was a monumental failure. Quaker sold it off three years later, and they took a $1.5 billion hit.

So the Gatorade acquisition was a poorly made decision combined with pure dumb luck. And the Snapple acquisition was a poorly made decision combined with no luck at all. And not even bad luck. Bad luck is what you get when you make a good decision, and it still turns out poorly. Right. That happens.

So these are the two factors that determine the outcome of a decision, the quality of the decision and luck. So if you want to improve your outcomes, these are the levers that you can manipulate. You can improve your decision making, or you can improve your luck. So let's start with luck.

Just kidding. Actually I'm not kidding. I'm going to tell you about one of my favorite bits of research.

Dr. Richard Wiseman recruited participants who self-identified as either consistently lucky or consistently unlucky. And then he gave each participant a newspaper, and he asked them to count the photographs in it. The unlucky people finished in about two minutes, on average. The lucky people finished in mere seconds.

On the second page of the newspaper there was a ginormous ad, and it read, "Stop Counting. There are 43 photographs in this newspaper."

[LAUGHTER]

The lucky people saw it, and the unlucky people missed it. Now, Dr. Wiseman did a bunch of research into the characteristics of lucky people. And what lucky people consistently do differently than unlucky people. And I'm not going to actually talk about it here, because in order to improve your outcomes, improving your decision making is a lot more effective, and, frankly, a lot more satisfying.

So I want to go back to the anti-pattern from earlier. This is a common failure mode in decision making. It's a narrow focus. When we're faced with a decision, we sometimes miss the big picture. Some of us fail to realize that there even is one.

A decision masquerading as a yes or no choice is a very common symptom of this anti-pattern. When making a consequential decision, we need more than one alternative. If you can get to three, you're doing basically better than most people. They need to be real alternatives. This is not the thing where you make up two bad alternatives just to get the three, or just to make the third look better. Right. This is not the thing where you're trying to trick other people's irrational brains. You're trying to untrick your own, which means that you need meaningful options. And they need to be meaningfully different from each other.

So this is easier said than done. Just saying, think of more options kind of doesn't work. Because we have this comfortable tendency to remain in our ruts.

And there are a multitude of ways that you can break out of a narrow focus. And I've come across three that have been particularly effective for me. The first of these is to take implicit opportunity cost and make it explicit.

Now, according to Dr. Shane Frederick, this may be as simple as reminding yourself that you have other options. In a number of his studies, they would present participants with a yes or no choice, whether or not to buy a movie, a pair of sunglasses, a dinner event.

The participants would be presented with the scenario and the price. And then half of the participants would be asked to evaluate a choice, to buy or not to buy. And then the other half of the participants were shown a variation of the not buy option that explicitly formulated the opportunity cost in terms of keeping the money for other purchases.

In the first group, 75% chose to buy. In the second group, 55% chose to buy. So when the opportunity cost was implicit, only 25% chose not to buy. And making it explicit, reminding people that they have other options, almost doubled that number.

Now, this is a bit nebulous. Right. Maybe the CEO of Quaker would have benefited from a gentle reminder that he could have spent the $1.8 billion on other things. But just reminding someone that they have other options doesn't actually generate those options. For that, you need a variation on opportunity costs that I like to call opportunity cost, dash, dash, force.

[LAUGHTER]

Assume for a moment that none of the alternatives that you're considering are available to you. What else might you do? If you're offered a job and you're wondering if you should stay at your current gig or take the new job, assume for a moment that you can't stay at your current job, and you can't take the new one either. What else might you do?

The second tactic that I've found useful for generating more alternatives is a variation on brainstorming. Now, our brains really like brainstorming. It's actionable. It's satisfying. It's ineffective in a narrow frame.

You can trick your brain by making it feel like it's brainstorming. But instead of exploring the solution space, you explore the problem space by brainstorming questions. And questions tend to spark more questions. And very quickly you'll get to dozens of questions. And then you can double the number of questions by taking every yes or no question and reformulate it into an open ended question. And then take the open ended ones and reformulate it into a yes or no question.

And sometimes one version of the question will tilt your brain in a more useful direction. And that'll generate more questions. At the end of it, you sift through and find the interesting questions. And these will often expand your focus even without answering them.

The third tactic that I've found useful for breaking out of a narrow frame is disagreement. Some people love a good disagreement. You might be one of them. And it's entertaining. You get this rush of endorphins. Sometimes you might enjoy the fight. Sometimes it's the winning.

Well, there's some satisfaction in pushing people's buttons. Or it might just feel good to sound smart and look rational.

I find disagreements really uncomfortable. My instinct is to avoid the at all costs. But I've found that disagreements are incredibly valuable. They're a shortcut to uncovering crucial information that you don't have. They're a powerful way of unearthing new options because they force you to discover new perspectives.

Just because they're valuable, doesn't mean that you want to go looking for them. Instead, just notice. Pay attention when your gut insists that someone is wrong. The painful and uncomfortable truth is, they are almost certainly right about something. And before you discount someone's perspective as irrelevant or wrong, try to understand where they're coming from. Make sure that you're understanding of their position is actually their position. Ask clarifying questions. Assume that they are rational, and smart, and then ask questions until you can tell how their point of view makes sense.

Now, you might still disagree. But you'll have learned something. So taking implicit opportunity cost and making it explicit, exploring the problem space with question storming, and optimizing disagreements for learning, these all help make individual decisions better. But they don't really help you decide how to decide when you need to make a series of related decisions over and over again.

Going back to the GitHub example, how do you decide which events to sponsor and which ones not to sponsor? What does a sponsorship strategy even look like? What is strategy?

Well, it's not a litany of decision trees and tasks. A strategy contains a unifying insight. It's not an inventory of hopes and dreams. A strategy is firmly grounded in reality.

It's not a concoction of fancy sounding words, creating the illusion of high level thinking. Strategy is a plan for overcoming obstacles. This means that you have to actually identify the obstacles. This is where people tend to fall short. They skip right past the part where they figure out what challenge they're facing, and what strengths and opportunities they have that provide leverage against the challenge.

So I'm going to give you two real life examples of what a plan for overcoming obstacles can look like. During the Cold War defense planning in the United States boiled down to estimating Soviet weapons inventory, and then making a shopping list. And then budget would be allocated. Never as much as they wanted. Purchases would be made off the top of the list. And then it would all start over again the next year for the next budget cycle, over and over again, year after year.

And then in 1976, a couple of folks in the US Department of Defense pointed out that the Soviet Union was strapped for cash, and they suggested that the United States should invest in technology that would force the Russians to spend a lot of money in ways that wouldn't make them more dangerous. This made it a lot easier for the Americans to decide how to allocate their resources, what to focus on, what not to do. And 15 years later, the Soviet Union was broke. And the Cold War wound down. And my dad moved to Russia.

[LAUGHTER]

So the other example is from the movie industry. In 1986 Ted Turner purchased two movie studios, Metro-Goldwyn-Mayer, MGM, and United Artists. Now, Turner had absolutely no interest in producing movies or dealing with film processing, or distributing television shows. So he sold most of it off piecemeal three months after making the acquisition, mostly back to the people he had bought it from.

And he kept one piece. And that was the television and movie rights to the entire catalog of classic films owned by these two companies. Turner had identified something known as the Lindy effect, which states that for every day that something survives, the chance that it will continue to survive increases. In other words, classic movies stay classic, and they become more classic over time.

This catalog was worth about $100 million when Turner purchased it. And then he launched Turner Network Television and Turner Classic Movies, where all of these movies were playing nonstop. And he created a classic movie empire that was worth billions.

So both of these examples of plans to overcome obstacles contained three key features. The first is the diagnosis. A diagnosis defines the challenge that they're facing, and identifies the opportunity that's inherent in that challenge. Turner identified that the Lindy effect was in play and that there was this catalog of classic movies that had untapped potential. And the US identified the juxtaposition of United States wealth and the Soviet Union's lack of resources as a situation which could be leveraged.

The guiding policy defines, in broad strokes, how you're going to deal with the challenge. So Turner was going to exploit this library of classic films. The US was going to make the Russians spend a lot of money in ways that wouldn't make them more dangerous.

Coherent action is where you devise concrete plans and feasible steps that all align with the guiding policy. Turner started TCM and TNT. The US did things like making their submarines quieter so that the Russians had to invest technology in better detection.

So strategy is, at its heart, very, very simple. A good strategy can be described on the back of a napkin. And in hindsight, it seems obvious.

So going back to the problem of sponsorships at GitHub, I eventually realized that my mandate was wrong. It makes no sense to have a sponsorship strategy. A sponsorship needs to be a tactic in the service of a strategy. All of these requests were an obstacle, but they weren't a strategic obstacle. We could successfully sponsor events without ever successfully reaching any of our goals.

So the recommendation that I brought back to upper management was that we flip the decision on its head. Instead of making a yes or no call on an individual basis within some global sponsorship budget, each team needed to look at what they're trying to achieve, and then figure out whether or not sponsorships would help them do that. And if so, they can negotiate a budget specifically for those sponsorships.

So sponsorships are not entirely altruistic. They're also not entirely self-serving. GitHub has a vested interest in helping the multitude of developer communities thrive.

I work on the ecosystem API team at GitHub. And almost everything we're working on is about improving the GitHub platform in such a way that third parties can build more interesting, more powerful integrations, and that they can do it with less friction.

One of the technologies that we've invested quite a lot in in the past few years is GraphQL. We want people in the GraphQL to be building tools, and developing expertise, and spreading interest. So now we have a budget for things like sponsoring GraphQL events.

I used to think that strategy was not for me. There's a saying, you can't see the forest for the trees. A friend of mine once said, Katrina, you can't even see the trees cause you're in the weeds. And it's true. Strategy seemed inaccessible, beyond my capabilities, beyond my paygrade.

Learning to make better decisions, or any decisions at all, that was powerful. But realizing that strategy is not magic, strategy is learnable, that was a revelation. And if there's anything I want you to take away from this talk, it's that strategy is not only for the people in the boardrooms. It's not just the domain of the execs and the VCs. We can all do the legwork to come up with a good diagnosis. We can all define in broad steps how to deal with the challenges that we identify. We can all devise feasible steps and coherent plans. We can all produce a plan for overcoming obstacles.

Is it easy? No. It's a lot of hard work. It gets easier with practice. And if we do that work, the result is that we put decision making into a framework. Strategy is a powerful multiplier for decision making. Each decision becomes less costly, worth more.

And it'll ensure that the efforts that you put forth are more effective, and the time that you spend is more meaningful, and that you don't accidentally move to Russia.

[LAUGHTER]

Thank you.

[CLAPPING]