Levels of UncertaintyBy CIOinsight | Posted 09-23-2002
Strategy-Making in Uncertain Times
Dr. Hugh Courtney is a teaching fellow at the Robert H. Smith School of Business at the University of Maryland and a strategy consultant to corporations. Now affiliated with Decision Strategies International Inc. in West Conshohocken, Pa., he is a former McKinsey & Company consultant and the author of 20/20 Foresight, which offers an approach to making strategy amid uncertainty.
CIO Insight: Has Sept. 11 changed the way companies do strategy?
Hugh Courtney: Yes and no. First of all, business strategists didn't face a certain world on Sept. 10 and then an uncertain world on Sept. 12. In fact, business strategists always face fundamental uncertainties, including those of national security and the effects those have on the macroeconomy. But I do think that because Sept. 11 was such a horrific, shocking event here in the U.S. and around the world, it certainly has created a greater awareness and perception of the uncertainties that businesses face.
What we're seeing is two reactions to this increase in perceived uncertainty since Sept. 11. One is a fortunate and very positive reaction, and one is a bit unfortunate. The unfortunate one is a sense that the world is so uncertain, so scary, the uncertainty is so impenetrable that we're seeing a bit of almost withdrawal from strategic analysis: Why do any of it, because you can't predict the future anyway? It's a waste of time. And this also leads to decision paralysis, where companies and individuals are afraid to make commitments because of all the uncertainties that they perceive.
A second, and I would say more fortunate response, and I think this is a majority response within corporate America today, at least from what I see, is just a much more systematic, serious approach to uncertainty when making future decisions, much more emphasis on trying to analyze what can and can't be known and then tailoring strategy to deal with uncertainties that no one could know ahead of time and that even the most prescient forecaster just can't get right.
This more serious approach, I think, is very positive, because the traditional, stereotypical or prototypical strategic planning toolkit in place in most companies doesn't really address uncertainty very well at all, and in some ways actually almost pretends that the world is forecastable. And that if you do a Five Forces Analysis of the industry, you'll know exactly how it's going to play out, and all you have to do is figure out what the most profitable position is and make your commitment.
As I said, the fortunate response to the uncertainties of Sept. 11 is that people are getting much more serious about looking at tools that better address the reality of uncertainty in their business environments, tools like scenario planning, real options analysis, game theory and so on.
CIO Insight: We've been talking about what you've observed them doing. What do you recommend they do now?
Courtney: First and foremostand this seems quite simplisticis not to ignore the uncertainty they face and to take it quite seriously. It all starts with diagnosing the individual uncertainty they face.
What you find is that for most decisions that businesses face, there are a handful of very important value drivers, whether it be the cost performance with a new technology, the response of a competitor to a new pricing or marketing campaign and so on. But it's usually a handful of five, six or seven variables that really drive whether the strategy pays off or not, and what we have found is that if you can systematically start with those value drivers and get the best competitive and market intelligence out there, that will tell you something about the uncertainty you face.
Sometimes these variables, or at least a subset of them, can be reasonably predicted. It doesn't mean you're going to be 100 percent right, but the range of what's going to be wrong isn't going to matter for making the choice. Sometimes there are really only a limited number of ways that it can play out, a limited set of scenarios. Competitors will match your pricing policy or they won't, for instance, or they'll meet you halfway because that's what they've done in the past.
Sometimes a very broad range of outcomes could occur. For example, new technology could give you a cost performance advantage of anywhere from 5 percent to 25 percent, but it's just too early to know where within the range you're going to fall. And then, finally, there are some variables where really you do know next to nothing. I call these Level 4 uncertainties, and you'd be fooling yourself if you even thought you knew.
What we find is that if companies are very systematic about identifying those value drivers and then taking each in turn in trying to figure out what can be known and can't be known about them, they will have a very clear sense of the range of potential outcomes, a much clearer sense anyway. They will know what it would mean to actually hedge their bets, if that's something that they can afford and is desirable. They will know which scenarios they can influence and shape and which ones they can't and that maybe they have to hedge against others.
Again, it sounds simplistic, but if companies use the information that's out there, if they are systematic about it, we find that they can do a lot better job than winging it or using old-fashioned strategy tools that pretend that the whole world is reasonably predictable.
CIO Insight: And this approach applies to the ways the world has changed since Sept. 11?
Courtney: Oh, absolutely. For instance, let's take an example of a U.S. airline today. It is trying to craft origin-destination strategies and rethink the hub and spoke network, which a lot of them are doing today. If the airline thinks about its value drivers, a big driver is business and leisure demand for travel on different origin and destination pairs, and what people are willing to trade off, such as price for convenience. The macroeconomic spillover effects from Sept. 11 and what's happened with consumer confidence in terms of the demand for secure travel have an enormous impact on whether or not those strategies will play out. And those variables are definitely more uncertain today than they were pre-Sept. 11 for those companies. In fact, they have probably a much broader range of possible scenarios for how some of these value drivers might play out over the next five or ten years than they thought they did on September 10.
And what I would recommend to them is not to go to the extreme of saying, "Well, it's a completely new world, we can't know anything about those value drivers today," or, on the other hand, saying, "Well, we'll use the controlled standard model approaches we did pre-Sept. 11." Rather, I would say the residual uncertainty is undoubtedly much greater, and you need to think about new scenarios. But that doesn't mean that you just have to throw up your hands and say none of this can be analyzed systematically.
Levels of Uncertainty
Levels of Uncertainty
CIO Insight: Your approach is to break down uncertainties into different levels, and then apply certain tools. Can you describe these tools?
Courtney: It depends on the type and level of uncertainty face. Do you face a discrete set of possible outcomes? Examples would be if the primary source of uncertainty is whether a competitor will match your price or build a plant or not, or whether a pending regulatory or legislative policy will take effect, or whether one technology standard will replace another.
The types of tools that tend to work quite well here are a class of tools or decision analysis tools that are about probabilistic decision-makingtrying to understand, even if you can't precisely quantify them, the relative probabilities, the relative likelihoods of the different outcomes, then assessing strategies against those potential outcomes. Which sorts of strategies, for instance, are robust, will pay off regardless of scenario? Which ones are more like big bets on one scenario or another, and what can you do to hedge those bets? Some traditional planning tools work well. Game theory is particularly helpful when you face competitive conduct uncertainty, when one of those uncertainties is how a competitor or even a supplier or customer will react.
In other types of uncertainty, which I call Level 2, you can only bound a very broad range of outcomes. Scenario planning has been used successfully for that type of broad range of uncertainty. However, what it can't do is place probabilities on outcomes. It can't do quantitative risk return analyses. Instead, scenario planners tend to use the range of potential scenarios to qualitatively test robustness, test what strategies are big bets and what's the worst potential outcome. Level 3 tools are quite similar but used in different ways. And a short way to think about the difference is that Level 2 can be more quantitative. In level 3 it starts to get a bit more qualitative, yet still quite systematic.
And then, finally, Level 4 situations are those where no matter how good your analysis, how thorough it is, there's just no way that one can bound the range of outcomes. So scenarios don't make sense, because who knows whether the scenarios are representative at all or whether they cover the range of outcomes? What most strategies do is work forward from analysis to conclusion. You say, OK, here's what the industry looks like. In fact, it will support a price point of $5 per widget, and we have a cost structure that is $3 per widget. And so we can make $2 per widget, so it's worth investing $10 million in that market.
In Level 4 situations, you don't know the price of widgets, and you can't even bound the range of outcomes. You don't know what the competitive response will be, and again, there really is no good analysis you can do to bound it. Instead you have to just work backwards from the potential solution to at least a set of assumptions. For instance, you want to invest $25 million in an unproven genomics area, where no matter how good your analysis is, you're just not going to be able to bound the commercial applicability of that area. You just don't know yet.
What you could do, though, is work backwards and say, OK, to support a $25 million investment today, you would have to believe that a commercially viable drug will get to market within X number of years, that it will have a potential patient treatment size of Y people, that the marketing dollars to bring the drug to bear will be less than or equal to Z dollars, etc. In other words, develop a scenario that would say this is what you would have to believe at the very least to support that $25 million investment.
Then at the very least what you could do is test industry analogies, case studies, etc., to try to see, for instance, whether this has to be the most successful drug launch ever in history to support a $25 million investment. At the very least you could, first of all, be very explicit about the assumptions, bring to bear some information from analogy and other case studies to assess those assumptions, and then make a more informed decision.
So there's no way that analysis can tell you that you're going to make 25 percent return on assets on the investment, nor will it even be able to bound the range of outcomes. But it would at least allow you to systematically assess the underlying assumptions that are implicit in the strategies you're looking at, and thus allow for more informed decisions.
CIO Insight:Of all the factors that have to be taken into account in the strategic planthe competitors, customers, markets, whateveris there any one of those that is more affected by Sept. 11 than the others?
Courtney:: Certainly I think geography is an issue, and for a multinational corporation in particular location security within that location is increasingly important. For instance, people may think twice or more than twice about investments in Indonesia today relative to what they were thinking about a couple of years ago, or anything in the Southeast Asia area. And, of course, obviously, the Middle East.
But I think the security issue first and foremost is around employees, then, of course, assets. You don't want plants being blown up. But it's actually much more subtle in that more and more people are thinking about institutional security, different political and economic regimes in different areas of the world. They're thinking more about security of their value chains, supply chains so that, OK, we're in a safe country, but actually all of our material comes from Indonesia. How do I think about that?
So the greater macropolitical economy issues are ones that more and more companies are addressing today, along with the fundamental economic institutions, the legal structures, the threat of political and social upheaval and unrest.
CIO Insight:What is the first priority for a company trying to make strategy in this era?
Courtney:: Well, the first priority is to calm down, forget the hype and do what you've always done best, which is to systematically understand what creates and captures value in your industry. And most companies will find that the majority of those value drivers are still the important ones, and if it was important to be cost leader in your industry pre-Sept. 11, it certainly still is, and you better focus on that. If differentiating product and service strategies are what drive your niche, then that's probably still the case, and so on. You shouldn't lose focus of those things.
On top of that, for some companies, some industries, there's going to be significant new uncertainties that they're going to have to address. But I would say they don't require a crystal ball, they don't require information that's not potentially available out there. It's just that you have to be systematically thoughtful about it.
And I know it sounds almost sterile, but the fact of the matter is we all felt an incredibly personal kind of horrific shock felt after Sept. 11, and some companies are facing almost horrific challenges today, but systematic, thoughtful management and leadership is the answer now more than ever.