McKinsey Quarterly

Decision making in your organization: Cutting through the clutter

| Podcast

Managing the decision-making process in a company is a crucial part of maintaining a well-functioning organization—which is why much more attention needs to be directed at how decisions are being made. In this episode of the McKinsey Podcast, senior partner Aaron De Smet and senior expert Leigh Weiss speak with McKinsey’s Simon London about how organizations can make better decisions and execute on them in a clearer, more efficient way.

Podcast transcript

Simon London: Hello, and welcome to the McKinsey Podcast with me, Simon London. As the existential philosophers observed, you are the sum of all your choices. If this is true of individuals, it’s doubly true of organizations. The choices made by managers, the decisions they take, well, they pretty much determine the fate of companies.

So, how do we improve the quality of decisions made by managers? How do we raise the overall quality of decision making? Joining me today to discuss all this are Aaron De Smet, a McKinsey partner based in Houston, and Leigh Weiss, a senior expert based in Boston. So, Leigh and Aaron, thanks very much for being here.

Leigh Weiss: Happy to be here.

Aaron De Smet: My pleasure. Glad to be here with you.

Simon London: Let’s start with when you’re working with clients, how do you know that this is a problem? How can you tell? What are some of the telltale signs that a company has a problem with its decision making?

Leigh Weiss: One of the things that I see most often is when you talk to employees about decisions you would expect them to make and they say, “I don’t actually make that decision.” And I’ll say, “Well, who’s your manager?” And they say, “Well, that person doesn’t make it either. It’s actually made by the CEO.” There is this real increased tendency for decisions to bubble way up to the top of the organization, either because employees don’t feel empowered to make decisions or because they are afraid to.

Want to subscribe to The McKinsey Podcast?

Aaron De Smet: The other telltale sign is when you have a meeting that’s meant to be a decision meeting, and they spend all their time just sharing information. Or you have a meeting that’s meant to be a decision meeting and they struggle to make the decision in the meeting, for various reasons. You start seeing the meeting before the meeting, and the meeting after the meeting, and the road shows to get everybody aligned because there’s a huge room of people and any one person seems to be able to grind the decision to a halt by asking a question or asking for more analysis.

A lot of these problems are not just problems of poor decision quality; they’re problems of slow decision making. In my experience, at least as big of a problem, and perhaps a much bigger problem, is decision velocity. The world’s moving fast, and you can’t afford to wait. Making a very great decision of high quality way too late in the game doesn’t help the company very much.

Leigh Weiss: We’re also talking about making the decision. I think there’s a really big problem we often see once the decision is made, which is that if people weren’t on the side of the decision that was made, they often don’t commit to it. So, you get a half-hearted execution of it.

Aaron De Smet: I sometimes tell my clients a joke: “Five frogs were on a log, and one decided to jump off; how many were left? The answer is five, because deciding to do something and actually doing it are two different things.” And this is true. Sometimes you think you’ve made the decision and you walk out of the room, but a month later, nothing’s happened.

This, again, is a telltale sign that something’s wrong in the decision-making process. The decision-making process should be a choice, where you have a level of commitment that drives action. If the commitment and action isn’t there, then something’s wrong in the decision process, itself.

Five frogs were on a log, and one decided to jump off; how many were left? The answer is five, because deciding to do something and actually doing it are two different things.

Simon London: What strikes me as interesting about this topic is that surely a lot of what managers do is make decisions. It’s certainly what executives do. We’ve had management and big organizations for well over one hundred years now. How is it that managers haven’t cracked this? It seems like such a core management discipline.

Leigh Weiss: I think that the context keeps changing. First of all, this is something that everybody needs to learn as they grow up through the ranks. On top of that, we have lots of things that are different from when managers were making decisions 20 years ago.

Our organizational structures are much more complex. I have a client that had a three-dimensional matrix: a function, a geography, and a product. It was enormously complex to make decisions. Some of the other issues are that we now have access to so much data and we do have increasingly good decision-making tools, but sometimes the discussion around the data is more important than the data itself. We also have a new generation of people, millennials, who are much less comfortable making decisions in hierarchies and want to grab the reins and move faster.

Aaron De Smet: The other thing that we’ve observed is some best practices around decision making are situational. For some types of decisions, those best practices work brilliantly, and for other types of decisions, they’re terrible. If you don’t apply the right best practices in the right way at the right time, you can get things that don’t work. It’s not enough to say, “I have experience, and I know what makes a good decision.” You have to say, “What am I optimizing for?” With decisions that can be quickly undone, you should take a lot more risk in making a wrong decision, because you can undo it. Decisions where the stakes are high and you can’t undo them need to be a lot more thoughtful and carefully planned.

Leigh Weiss: Another good example of that, Aaron, that I’ve seen is in some situations a decision can be made very effectively by one person. They may consult a few others, but they know enough to make a good decision quickly.

In other cases—often, cases where you’re making a major decision at the company, the kind of decision that might be about a merger, it might be about a new strategic direction—you want to have a lot more debate.

Simon London: Is it a good place to start if you’re working with a company on this to think about different types of decisions and the types of processes that should apply to them? Doing that kind of mapping?

Aaron De Smet: That is a good place to start. What you’ll find with some organizations is they’re very good at one type of decision and not very good at another type.

Simon London: Do you want to walk us through the typology here? How many types of decisions are there that one could sensibly bucket?

Leigh Weiss: Sure, we’ve found that it’s helpful to talk about four different kinds of decisions. One is your classic big-bet decision, where you’re making a decision that’s going to have enormous implications for the company. It’s often not easy to undo it. It might be an acquisition or a merger. It might be a major capital investment. That’s the first type.

The second type is a decision that isn’t actually a single decision. We call it a cross-organizational or a cross-functional decision, where many different parts of the organization are involved and there are lots of little decisions that accumulate to a larger decision. A good example of this might be something like pricing or decisions in a supply chain.

The third type of decision is one that can easily be delegated to a particular role—somebody who has enough knowledge to make a good decision, may interact with other people to get feedback and perspective on making the decision—but does not need to be made in a committee and does not need to be drawn out. It does not need a carefully mapped decision process.

Aaron De Smet: There is a fourth type, ad hoc decisions, which are decisions that are infrequent and reasonably small and contained, where you don’t try to figure it out or map it out ahead of time. You just say, well there’s a bunch of stuff that might bubble up, and we’ll deal with it as it comes up. We’ll cross that bridge when we come to it.

But the other three types that Leigh describes can and should be segmented ahead of time. What are the decisions that matter of the big-bet nature? There are all kinds of those. You might have read recently in the paper, Amazon has a contest to pick where its second headquarters will be. That’s a big decision. That has big implications for everybody. That’s a decision that should be treated like a big bet. It’s a discrete decision where there will be a point in time when the choice is made.

Simon London: Let’s take each of those in turn, certainly the first two. Let’s take the big bets first, and then we’ll do the sort of ongoing cross-cutting type things. You mentioned there are some best practices here for the big-bet decision. What would you point to? I’m sure there’s quite a few, but what are the three or four things that companies should do?

Aaron De Smet: This is the one where most of the organizations I work with are pretty good at it. Because they’re visible and big bets, senior executives and your best leaders tend to get involved, they tend to be good at it. The one area that sometimes they’re not as good at, which is one of the big best practices, is trying to make sure you aren’t biased in your decision-making.

Decision biases can creep in, especially as you’re debating over time. Biases can significantly distort decisions. And since, in this case, the quality of decision is paramount, a little bit of decision bias can lead to a really, really bad decision.

Let me give some examples of biases that can affect decision-making. One of my favorites is the endowment effect. The endowment effect happens when someone values something more just by the fact that they have it. There was some great research where a professor walks into his class, and he gives half the students in the class a mug. He asks how much it would be worth and what would the other students have to do to buy the mug from them. And the amount they charge, that they say it’s worth to them once they’ve been given the mug, is significantly more than what they would pay for it had they not already owned the mug.

They’ve tried this in all settings, with all different types of participants, with all different types of objects. The fact of owning something makes you value it more. When it comes to, let’s say, an activist investor saying, “You should break up your company,” that bias can definitely creep in. You think, this is worth way more than that. We should keep it.

Leigh Weiss: I would say there are a few other issues that are important also for big-bet decisions. One of them is that executives making big decisions need the right information. In many cases, they don’t have it.

There’s also the opposite problem, which is that they have all of the information, and they tend to focus just on the facts on pieces of paper or PowerPoint slides rather than the debate among the executives that’s really most important and most valuable for thinking about alternatives, what are the different options, how might we approach a big-bet decision. It is in that engagement where companies get to really good decisions.

The other thing that I’ve seen go wrong a lot, even with big-bet decisions, is where executives will make a decision in the room and then, if some of them are not happy with it, they’ll undermine it afterwards. An essential best practice is committing to a decision once it’s made. Amazon’s Jeff Bezos calls this disagree and commit. It’s something that can really slow an organization down if they don’t do it.

Another cognitive bias, Aaron, that we see happening a lot is confirmation bias, where people think they know what the answer is and so they read the data and information in a way that supports the answer that they want to get to, and they tend to ignore the arguments or pieces of information that would sway them away from making the decision that they want to make.

Simon London: The other thing you probably see at a senior level, and certainly a middle-management level, is you often don’t get clear communication of what the decision was, in writing, after the decision has been made. So, it’s just very, very clear and they have a record of this is what we agree, you’re going to do this, you’re going to do this, you’re going to do this. I’m guessing senior teams do this as a matter of course.

Leigh Weiss: Sometimes they do. I’ve often seen that organizations are not very good at communicating decisions once they’re made. It’s one of the reasons why you tend to get way too many people in meetings, because they think they need to be there for informational purposes. If they’re not there, they won’t hear what’s going on.

So, there should be very clear and standard practices about communicating decisions after the meeting. Not only that, it’s important to say what the decision is as well as what the implications are cascading through the affected employee groups.

Aaron De Smet: For big-bet decisions, most executive teams are good at communicating the decision. What a lot of the people who need to carry out those decisions want to know are two things in addition to the decision. Why? Because why gives them context. It gives them more clarity on how this connects to other things and what the full set of expectations are about what the decision is supposed to produce and why we made it and what the tradeoffs are. A lot of reasons why people want to be involved in a decision isn’t just to know what the decision is, it’s to understand why it was made and how it was made. The second is, what are the implications? So, in communicating the decision, not just here’s the decision we made, here’s why we made it and here are the implications for how it needs to be carried out.

The last thing that’s helpful for big-bet decisions, and this is one that I think sometimes doesn’t get addressed, is being clear on what the question is we’re trying to answer. What is the decision we’re really making? You can say we’re deciding whether to buy this company, yes or no. But often, that’s part of a bigger decision of inorganic growth. And there’s a much broader set of options for growth, both organic and inorganic, and we’re deciding whether to buy this company for reasons of a goal or an objective we have around inorganic growth.

Simon London: We’ve dealt with the big bets. I’m guessing the best practices for cross-cutting decisions are a little bit different. Why don’t you talk us through them?

Aaron De Smet: I think this is the trickiest one. This is the one where people most often want to simplify it to a discrete decision and apply the best practices of either a big-bet or a delegated decision. There are a few flags for me when people are treating this decision wrong.

If a decision is actually the culmination of many decisions in a process over time, different decision makers need to align. For example, pricing. The price that our customers pay for different products is not just decided. We have list price, we have floor price, we have exceptions, we have rebates, we have variations over time, we have sales managers who can make exceptions or escalate them. It’s not just decided one and done. That’s a good example that Leigh had mentioned before. The flag for me is when somebody says, “Well, who’s the ultimate decision maker?” There isn’t one. Who has the D on pricing? Well, nobody. There is not one person.

If you try to vest all of that decision-making authority for pricing in one single person, say the sales manager or the finance person or the marketing person, you often run into problems where it slows down the ability of the organization to make fast, effective decisions. You can vest decision authority at different points in the process, like who has the ability to set the floor price. Fine. But that’s not all of pricing. That’s just one decision as part of the process.

Simon London: So, it sounds like in these cases, it’s the choreography that’s going to be important. It’s who you bring into the conversation at what point. It’s the order in which decisions get made, laddering up to your price point.

Leigh Weiss: Right. And do people understand where the handoff points are, where the choke points tend to be, where it might be helpful to have an escalation process or bring people together to debate a particular issue?

I often find in mapping decisions that if I ask a client to do it, they’ll start out with 15 steps, and you can often reduce it to half of that.

Aaron De Smet: When I do this on cross-cutting decisions, I often start by mapping the process out from beginning to end. Then I take out all the activities that don’t fit into one of these three things:

One, it’s a hand-off. One group or person or department is handing something off to another group or person or a department.

Second, it’s a critical point of alignment where some group of people need to get together to agree or align on something or coordinate something.

Or three, it’s a really important decision. There’s a point in time where a decision gets made and we need to be very clear who makes it, whether that’s one person or whether it’s two or three people all agreeing on the decision.

If you strip out all the other stuff and are just left with a process that highlights those things, then you can work on the choreography. You can say, “Let’s trust people to do their jobs and get the activities right. We don’t need a super detailed activity map. What we need is a decision map.” The decision map can help greatly with the choreography.

Leigh Weiss: I’m reminded, listening to your example, of the work that you and I did a few years back with a consumer company, where we were looking at different decision processes. What we found was that 40 percent of all of the interactions that people had in these decision processes did not add any value. It wasn’t because the employees were dumb or not doing their work, it was because they didn’t need to be involved. It happened to be a culture that was highly inclusive. When we took out the unnecessary interactions, what happened was people were much happier because they could spend time on the things that mattered. The decisions flowed much faster, you got the velocity, and you got to better outcomes.

So, ungumming-up the works, which is what you’re describing, is a very useful way to do that. I often find in mapping decisions that if I ask a client to do it, they’ll start out with 15 steps, and you can often reduce it to something like half of that.

Simon London: An interesting layperson’s question here is, does there need to be a role or a group of people who are specialists at this kind of decision-making hygiene?

If there’s a chief technology officer, I can imagine that part of the office of the CTO is to make sure that these kinds of technology decisions are really thought through and debated and discussed. But for a lot of decisions, does there need to be a chief decision officer or a group of people who specialize in making sure that the quality of decisions and decision-making processes across the organization is high?

Leigh Weiss: You’ll often see a chief of staff–kind of role whose job will be to make sure that the right information is there, that the decision to be made is clear at the beginning of the meeting, to make sure that the participants in the meeting are all engaged and bringing their views to the discussion, and then, after the meeting, to make sure that the decision is communicated and executed.

Aaron De Smet: For a decision meeting, and this could be for your big-bet decision or for those critical alignment points in a cross-cutting process, it is often helpful to have someone in charge of the staff work. It doesn’t have to be a chief of staff per se, but it’s someone who’s in charge of managing the meeting.

I don’t think you want one person in the organization to be your decision master who figures all this stuff out. Actually, I think you want most of your senior leaders to be good at decision making and thinking it through. Having said that, it is helpful sometimes to have a meeting manager of some sort who’s helping with the staff work, who helps make sure the agendas are tight, who helps ensure that in a decision-making forum, most of the time is spent debating and deciding and not just sharing PowerPoints. So, it does help.

Simon London: In many cases, it’s the functional leader, to my point about the chief strategy officer or the CTO. Your CMO [chief marketing officer] or your COO, your head of supply chain, really should be the owners of the process, even if they’re not making every decision along the way.

Aaron De Smet: Yes, although it’s not always clear. So, let’s take sales and operations planning. This is one of the supply-chain processes that really connects procurement and manufacturing, buying and making of the stuff, all the way through the selling and delivering of the stuff. It’s not always obvious who should run that. Should it be an operations person? Should it be a manufacturing person? Should it be a salesperson? In my experience, it’s often helpful for it to be a supply-chain person who’s not, strictly speaking, manufacturing, but it doesn’t have to be. It could be a commercial person. It could be someone in charge of a product line, since supply chain is usually product specific.

Leigh Weiss: Whoever it is, they need to have the perspective of the multiple functions that go into it as well the operations, the line issues.

My experience is that these individuals Aaron is talking about are often very good at identifying where a problem is. They can say, “Okay, we’ve got a big quality problem in the production line in this particular part of the world.” Where I think we need a lot of improvement across this kind of a role is in people being able to be better diagnosticians to understand, well, why do we have a quality problem? What is it about? And break it down into the component pieces.

Untangling your organization’s decision making

Untangling your organization’s decision making

Then, because they don’t have ownership over the whole thing—because by nature it’s cross organizational, cross functional—they need to have very strong influencing skills. I would think a lot, if I were leading an organization, about the capabilities that these people in highly cross-functional roles for pivotal processes and decisions have.

Simon London: Everything you’re describing sounds very process driven. It sounds very complicated. It sounds the opposite of speedy and agile. I know speed and agility are something you mentioned earlier as something companies need to work on. So, isn’t this all a little bit sclerotic?

Aaron De Smet: No. No, it’s not, for three reasons. The first is we haven’t talked a lot about delegated decisions. And that’s where you can really get speed. If you can take decisions that are neither the big-bet nor the cross-cutting that can be delegated—and actually delegate—you can delegate to leaders or natural working teams and empower them to go fast.

One of the trickiest foundational things to get in place is the decision architecture.

The second is for things that aren’t those simple delegated decisions, getting the process right is huge because this will allow you to move quickly. How many times have you been in a room where you thought you made a decision, only to find out there was somebody else who wasn’t in the room or who escalated to their boss that they weren’t really happy with it, and the decision gets reopened?

Decision churn is a huge problem in a lot of big, complex companies. Most of the organizations I work with are large, global companies. They’re complicated. And decision churn is a huge problem, where people don’t know when the decision is finally made. Even when they’re in a room and they think they made it, they have to have the meeting after the meeting to go see if it’s really made or if somebody’s going to go throw a veto out there and reopen the whole thing.

Finally, I would say, the research we’ve done recently, and there was recently an article published about this, says there are a bunch of things you have to do to be agile. A lot of the things to be agile are kind of sexy, advanced, fast, dynamic, cool things to just move at much, much faster velocity. But there’s a set of things that are very foundational. If the foundations aren’t in place, you might be able to do some cool, fast, agile pilots, but you will never scale it up if you don’t have the foundation.

One of the trickiest foundational things to get in place is this decision architecture. Without a decision architecture that works and that has some level of process and has some level of discipline and has some level of standardization, the rest won’t work. It just becomes chaos, or it invites a different kind of bureaucracy.

Simon London: We haven’t spent much time talking about the delegated decisions, that third bucket. What do companies often get wrong in this area?

Leigh Weiss: I think they get two things wrong. One is that they often are not clear about who has been delegated to have the decision rights. In one financial-services organization that I worked with, the senior executive, at the beginning of every meeting, announced who was going to make the decision and where the decision was delegated.

I think the second problem is that often, people do know that they’re the delegated decision maker but for a couple of reasons they don’t end up making the decision. One is that they don’t feel empowered to make it. A second is that they feel afraid to make it—if they get it wrong, they’ll be punished. A third is that they’re not allowed to make it because somebody senior to them intervenes or pulls it back.

Aaron De Smet: Yeah, this is the biggest problem—the delegated decisions are kind-of delegated; they’re not fully delegated. So, somebody is told, “You make the decision.” But over time, what they learn that really means is, “You make the decision as long as it’s a good one and it’s the same one I would have made. If it’s an important decision that I care about and I would have made a different decision, you’re going to be in trouble or I’m going to reopen it. I’m going to take the decision back.”

Then, what people start doing is they start saying, “If I really know exactly what they would do, I’ll go ahead and make it, otherwise maybe I’ll escalate it and check in.” And you have this problem, this escalation problem. Sometimes, it goes up multiple levels. So, even though it was meant to be delegated, it just bubbles right back up because it was only kind-of delegated.

Leigh Weiss: This ends up having huge implications for the culture of an organization because everything we know from the research on organizational health is that empowered employees are more satisfied, and organizations are better performing when employees are more empowered. But if they don’t feel like they can be, for the reasons we’ve been talking about, you lose all that benefit.

Simon London: Okay, so that’s all we have time for. Thank you very much, Aaron De Smet and Leigh Weiss. Thanks for being here today.

Leigh Weiss: Thank you.

Aaron De Smet: Thanks.

Simon London: To learn more about our work on decision making, organizational health, organization agility, all that good stuff, please visit us at

Explore a career with us