January 25, 2010

McKinsey, Scenarios and Us

By Patrick Marren

The November 2009 issue of the McKinsey Quarterly includes an article by Charles Roxburgh entitled “The Use and Abuse of Scenarios.” It includes a number of good tips about scenario-based strategic planning, based on his experience in building scenarios over the past 25 years. It also highlights some important distinctions between his understanding of scenarios, and the way in which FSG has gone about creating and using them over the past few decades. And finally, it brings to the surface the urgent concerns of executives as they go about leading their organizations under uncertain conditions.

Mr. Roxburgh says a number of things that FSG agrees with wholeheartedly:

  • “Scenarios expand your thinking.”
  • “Scenarios protect against ‘groupthink.’”
  • “Scenarios allow people to challenge conventional wisdom.”
  • “Don’t rely on an excessively narrow set of outcomes.”
  • “Don’t chop the tails off the distribution.”
  • “Don’t use a single variable (e.g. “Best/Medium/Worst).”

All of these points essentially boil down to his first: the purpose of scenarios is to expand one’s thinking. Even the “don’ts” are meant to keep scenario users from narrowing thought.

At a certain point, though, Mr. Roxburgh begins to alternate between this expansive point of view and a seemingly countervailing urge to narrow focus and cut off thinking. And while this urge seems to go against the whole grain of scenarios as we understand them, it does reflect the very real pressures toward concreteness and pragmatism and certainty that are felt by leaders of organizations. We still disagree with his prescriptions in some of these cases, but we understand where his prescriptions come from.

The first example is his warning that scenarios can paralyze organizational leadership – i.e. present them with a range of possibilities, and they might not know what to do.

In our experience, presenting an organization with multiple scenarios never causes organizational paralysis, because we never simply present leaders with scenarios and walk away. Scenarios are simply a tool to provoke strategic insight. (Sometimes that strategic insight can cause paralysis later on, as it challenges long-held assumptions and modes of thought, but that is another story.) Be that as it may, his cure for such paralysis is “is to pick the scenario whose outcome seems most likely and to base a plan upon that scenario.”

But a scenario is a detailed description of a potential future – it must be detailed if it is to be of any use for planning purposes. Yet as it becomes more detailed, the probability of it “coming true” approaches zero. The fact is that any particular detailed scenario, no matter how likely it seems, has a vanishingly low chance of actually occurring. Fortunately, scenarios are merely a planning tool to elicit creative strategic insights about what the future might bring. If our clients finish an engagement with the idea that there is or could possibly be such a thing as a “most likely scenario,” we will have failed them.

He goes on to say, “Companies can’t build all possible events into their scenarios and should not spend too much time on the low–probability ones.” But this seems to assume that there is such a thing as a “high–probability event.” Other than the sun coming up in the morning, it’s difficult to identify one of these in real life. Scenarios exist to challenge our thinking about what is “high–probability.” And if we are not supposed to “cut the tails off the distribution,” for fear of missing something significant that most people consider unlikely, then why on earth should we consign something considered “low-probability” to the dustbin?

Again, however, we acknowledge the pressure felt in C-suites to be practical and hard-headed. In the short term, actions must be taken, and while there is no such thing as certainty about exactly what will be, a leader wants to be certain at least that he or she is doing the thing most likely to be conducive to success in the short term, given all available information. In our experience, however, scenarios increase confidence even in short-term decisions, by rehearsing a wide variety of futures.

He then advises, “Don’t discard scenarios too quickly;” says that scenarios ought to be revised often; and recommends that one scenario should be replaced with a new one from time to time.

But individual scenarios are of extremely limited use; it is only as a complementary set that scenarios can do their job, which is to present the client with the fullest practicable range of plausible futures that could face them. The set of scenarios is carefully constructed to ensure that full range. Pulling a scenario that has been specifically designed to counterbalance the others, and throwing in another of completely different provenance, may skew and actually narrow the client’s strategic vision, not to mention undermine leadership’s confidence in the strategic reliability of the scenario set.

Mr. Roxburgh then urges us to “Bear in mind the one instance in which strategists will not want to use scenarios: when uncertainty is so great that they cannot be built reliably at any level of detail.”

Actually, we tend to think the opposite is true. It is precisely when uncertainty seems greatest that scenarios provide the greatest benefit, because it is then most of all that the “Great Blankness” of the future, that failure of imagination caused by surprise and unfamiliarity, demands the sort of analytical framework that scenarios can provide. More to the point, what’s the alternative? To stick to standard extrapolative spreadsheets full of misleading certainty? That cannot be the answer.

Mr. Roxburgh recommends that we “always develop four scenarios or more, never three.” A good recommendation; but when we dig a little deeper, we find another difference between our philosophy and his. The problem with three is not precisely, as Mr. Roxburgh states, that “people always pick the middle one.” The problem is that people will gravitate what they perceive to be the “middle” one, when in fact there ought to be no “middle” in scenarios.

But Mr. Roxburgh recommends a “base or central case” for another reason:

…[T]hose who must weigh scenarios and reach decisions based on them expect and deserve to get a specific point of view about the future. The scenario that is highest in probability should always be identified, and that ought to become the base case.

Mr. Roxburgh’s desire for “a specific point of view about the future” is well-taken. We definitely owe leaders this much. However, his solution – providing a scenario as a “specific point of view about the future” – puts the real benefits of the whole scenario effort in jeopardy. There is a far more powerful and surer point of view that emerges about the future from the entire scenario-based strategic planning experience. But it can only come from exploring an entire set of scenarios, creating strategies and garnering insights that have been stress-tested across a full range of plausible futures.

Building Strategic Confidence for CEOs
This point of view might bring to the surface previously unexamined critical assumptions; the hitherto-unappreciated criticality of a certain input to a product or service offering; the vulnerability of the organization to a certain type of competitor or a novel form of competition; or a strategic opportunity that was hidden in plain sight. Each of these could still vary widely depending on conditions; but through the experience of scenarios, the organization would have developed a sophisticated and dynamic understanding of them – one that a single scenario could never provide.

That is why it is vital that scenario planners not be forced into providing a “point of view” in the form of a scenario. That scenario will inevitably be taken as THE single point forecast of the future, and this goes against everything scenarios were invented to remedy.

Mr. Roxburgh exudes the sense of urgency and distaste for woolly-headedness that leaders of large organizations must possess if they are to be effective. But he may sell scenarios a little short in his article; and frankly, he may sell CEOs a little short as well. Scenarios, properly handled, should increase management confidence, not cause paralysis; and leaders of today are sophisticated enough, in our experience, not to require a false “most likely future” to which they can hold their subordinates. Scenarios can replace a false sense of security with a richer, more sophisticated and dynamic understanding of the external environment and the true range of plausible outcomes. And that should be the ultimate goal of any application of scenario planning, long-term or short-term.

Patrick Marren is an FSG principal.

Thoughts?