Mapping Strategy

Butterflies for Dummies

Check out this succinct review in last Sunday's Boston Globe of an important but oft-misunderstood concept critical to any forecasting or planning endeavor [emphasis and link added]:

The butterfly effect is a deceptively simple insight extracted from a complex modern field. As a low-profile assistant professor in MIT's department of meteorology in 1961, [the late Edward] Lorenz created an early computer program to simulate weather. One day he changed one of a dozen numbers representing atmospheric conditions, from .506127 to .506. That tiny alteration utterly transformed his long-term forecast, a point Lorenz amplified in his 1972 paper, "Predictability: Does the Flap of a Butterfly's Wings in Brazil Set Off a Tornado in Texas?"

In the paper, Lorenz claimed the large effects of tiny atmospheric events pose both a practical problem, by limiting long-term weather forecasts, and a philosophical one, by preventing us from isolating specific causes of later conditions... It is extremely hard to calculate such things with certainty... Realistically, we can't know. "It's impossible for humans to measure everything infinitely accurately," says Robert Devaney, a mathematics professor at Boston University. "And if you're off at all, the behavior of the solution could be completely off." When small imprecisions matter greatly, the world is radically unpredictable.

Moreover, Lorenz also discovered stricter limits on our knowledge, proving that even models of physical systems with a few precisely known variables, like a heated gas swirling in a box, can produce endlessly unpredictable and non-repeating effects.

"Lorenz went beyond the butterfly," says Kerry Emanuel, a professor in the department of earth, atmospheric, and planetary sciences at MIT. "To say that certain systems are not predictable, no matter how precise you make the initial conditions, is a profound statement." Instead of a vision of science in which any prediction is possible, as long as we have enough information, Lorenz's work suggested that our ability to analyze and predict the workings of the world is inherently limited.

What few articles on this subject touch upon, but I find endlessly fascinating, is how this undeniable, well-grounded scientific insight interacts with human nature -- specifically our ingrained need/desire to feel knowledgeable and in control. (I use the term 'we' in both the individual and corporate sense here.)

It has been my experience that few individuals and very few (if any) organizations ever err on the side of humility in this regard. Perceptual failures are seldom in the direction of assuming less than is actually predictable. More often, the response is something like this:

yes I know about the butterfly effect, chaos theory and all that... but it's kinda academic... and you don't know my boss... you see, I/he/she/they/we need to predict XYZ anyway...
because my/our future depends on it... so just give us a number... give us your best guess about the probability... please?

People fall back on a prediction/probability paradigm either because it's what they know (or have the planning tool-set to address) or because, even knowing that approach is flawed, they find it more comfortable... and comforting (or politically expedient in a particular organizational culture) to pretend that what is fundamentally uncertain is perhaps predictable after all. The chance of rain is 62.5%...

Thinking Exotically

With air travel on my mind after a West Coast trip last weekend, the following strikes me as positive, (and not just because I'm a marathon runner who tends to pack light). Rather, it's the kind of thinking we do routinely when we help clients develop scenarios. E.g., what if industry 'A' adopted the business model of industry 'B'? Emphasis added:

Imagine two scales at the airline ticket counter, one for your bags and one for you. The price of a ticket depends upon the weight of both.

That may not be so far-fetched.

"You listen to the airline CEOs, and nothing is beyond their imagination," said David Castelveter, a spokesman for the Air Transport Association, a Washington, D.C.-based trade group. "They have already begun to think exotically..."

With fuel costs almost tripling since 2000, now accounting for as much as 40 percent of operating expenses at some carriers, according to the ATA, airlines are cutting costs and raising revenue in ways that once were unthinkable...

Singapore Airlines... is "trying to eliminate unnecessary quantities of extra water" to save weight, Chief Executive Officer Chew Choon Seng said in an interview...

Robert Mann, head of R.W. Mann & Co., an aviation consultant based in Port Washington, New York [said,] "If you look at the air-freight business, that's the way they've always done it... We're getting treated like air freight when we travel by airlines, anyway."

This signals that an industry long characterized by minimal innovation starting to think beyond convention (but... we've always done it that way!) The shock is external and pressing in this case (fuel costs). It needn't be so. Better to do this kind of thinking every day.

Best-practice use of scenarios involves doing this systematically, comprehensively and, over time, making it an ingrained habit of strategic thinking among management. (How could we change the rules to our advantage?) Having several contingent strategies already fully 'baked' when a crisis hits can move one ahead of more flat-footed competition who must do all that thinking while under existential, immediate pressure.

Next up: refunds for in-flight diuretics.    ;-)

From Risk to Uncertainty

I don't agree with everything in this piece by Thomas Homer-Dixon that appeared last week in the Toronto Globe and Mail, but this quote is an absolute gem (emphasis added):

Our global financial system has become so staggeringly complex and opaque that we’ve moved from a world of risk to a world of uncertainty. In a world of risk, we can judge dangers and opportunities by using the best evidence at hand to estimate the probability of a particular outcome. But in a world of uncertainty, we can’t estimate probabilities, because we don’t have any clear basis for making such a judgment. In fact, we might not even know what the possible outcomes are. Surprises keep coming out of the blue, because we’re fundamentally ignorant of our own ignorance. We’re surrounded by unknown unknowns.

It's something I've said for a long time:

It's tempting to think that all things are predictable given enough information, enough minds, enough time and enough computing power. It's just not true. (Which is not to say that some things are not predictable... and with incredible precision... a phenomenon that leads to overestimating the scope of problems and questions that lend themselves to such methods.)

Telling which is which is the trick...

I would go even further to say that really smart people who, by life experience know that some things are fundamentally unpredictable still draw an unvoiced sense of emotional comfort in their business life from the idea that some wise expert somewhere has been to the future (for all intents and purposes) and if we could just find him or her things would be OK... and/or that a really sophisticated computer model or prediction market (the 'collective mind') can provide crystal ball-level insights.

Sometimes yes. Often, no.

I liken Mr. Homer-Dixon's observations to those tragically massive car pile-ups that happen a few times of year in fog-prone areas like the Central Valley of California. Everyone is driving along at a reasonable speed, with reasonable spacing between vehicles. People are sipping coffee, tuning radios, maybe talking on cell phones. Slightly distracted, but mostly responsible. All is normal.

Then the first guy hits a fog bank and can't see squat. He taps his brakes. The second guy sees red lights and fog coming up fast and taps his brakes just a little bit harder, and so on. In just a few seconds, hundreds of cars end up in a tangled heap and people die. All because the guy in front was convinced by every one of his senses and not without justification based on experience that the visibility on the next 100 yards of road would be the same as on the last 100 yards of road.

How Not to Map the Future

Not recommended in a corporate environment...
...or any other, for that matter    :)
[emphasis added]

Each future mapping system (tarot, astrology, tea leaf reading, I Ching, dowsing, palmistry, rune casting, etc.) has its own unique tools... Runes, Pendulums, Crystal Balls, Tarot Cards, Spirit Boards, Planetary Ephemeris, Psychic Cups

Scenarios as Vehicles for Fear-Mongering

When we develop scenarios with clients, we emphatically avoid the kind Frank Furedi (rightly) decries here [emphasis added]:

From global warming to obesity, bird flu to terrorism: 2007 was the year when the threat of an apocalypse became an everyday, even banal public issue... The fear market in apocalyptic scenarios continued to flourish in 2007. Almost every week we were told that ‘the situation’ is far worse than we originally thought... Public figures appear to have lost the capacity to reassure or lead people. Instead, they frequently opt for evoking frightening futuristic scenarios where the line between fiction and reality become unclear.

One consequence of Western societies’ obsessive preoccupation with the apocalypse-to-come is that less and less creative energy is devoted to confronting the all too important problems that exist in the here and now. Take the global credit crunch unleashed by the sub-prime home loan crisis this year for instance.

In terms of its material impact, this was arguably the most significant event of the year. After more than a decade of economic stability, the world economy faces the threat of a major recession with important implications for people’s lives. This threat may not make an exciting plot for a sci-fi movie, but it has a direct bearing on the quality of life of millions of people. It also raises important questions about an economic system that is so heavily reliant on using fictitious capital to reproduce itself.

Events over the past 12 months suggest that what we think and how we think influences how we experience our reality.

Some rules and questions we use to avoid these traps and test whether scenarios are useful include:

Are scenarios sufficiently orthogonal to and distinct from one another? Does each embody both 'good' and 'bad' elements? Real world developments are seldom all good or all bad at the same time and from the same perspective. If participants in a scenario workshop find it trivial to line up the scenarios in the same way from "good/easy for us" to "bad/frightening for us", we haven't done our job of representing real-world nuance in hypothetical future stories.

Do scenarios incorporate "here and now" events and choices? (We usually embody these in what we call 'events', a component of modular scenarios). Scenarios entirely about some far-off, visionary 'place' with no explicit ties to current issues are seldom useful beyond the fiction stacks.

Are scenarios directly comparable to current conventional wisdom? (I.e., as Furedi puts it, "how we think... [and] experience our [present] reality"). Without a concrete "you are here dot" scenario that represents what constituents are thinking and assuming, it's impossible to describe how "far away" hypothetical future scenarios really are, or what change they imply. If I'm contemplating a trip to Miami, it helps to know (in terms of budgeting, preparation and mode of transport whether I'm currently in Juneau, Ft. Lauderdale or Tiera del Fuego.

The Future Will Be... Different From What We Expect

Great insights from science writer Ed Regis [emphasis added]:

...after watching all those forecasts not come true, and in fact become falsified in a crashing, breathtaking manner, I began to question the entire business of making predictions...

...the source of this incredible mismatch between confident forecast and actual result [is that] the universe is a complex system in which countless causal chains are acting and interacting independently and simultaneously... all of them running in parallel, and each of them often affecting the course of the others...

It is therefore

...hopeless to try to specify in advance what's going to happen as they jointly work themselves out.  In the face of that complexity, it becomes difficult if not impossible to know with any assurance the future state of [a] system...

(other than tides, moon phases, etc.)

Further, it's an illusion to think that supercomputer modeling is up to the task of truly reliable crystal-ball gazing.  It isn't.  Witness the epidemiologists who predicted that last year's influenza season would be severe (in fact it was mild); the professional hurricane-forecasters whose models told them that the last two hurricane seasons would be monsters (whereas instead they were wimps).  Certain systems in nature, it seems, are computationally irreducible phenomena, meaning that there is no way of knowing the outcome short of waiting for it to happen.

After accepting this reality (something that's as or more true for systems in which human beings play a role, e.g., social, economic, political, business, etc.) the next best things to knowing and predicting are imagining and recognizing:

• imagining multiple, archetypal ways in which the future might evolve, and
  
• collectively being able to recognize (if only a little bit faster than the other guy with whom you are competing) the key events and sequences of events that signal (if only weakly) that one or more paradigms (and not others) are starting to unfold.

China's GDP Changes Hands Every Day...

Well, almost. In my business it has become habit to scoop up watershed factoids like this and--along with other bits and pieces--make sense of them in a larger context of multiple future scenarios.

...the Bank for International Settlements in Basel, Switzerland... [tracks] the size of foreign-exchange markets. In April, daily turnover in currency markets rose to $3.2 trillion, the bank said yesterday.

That's more in value than the annual economic output of Germany or China, changing hands in currency markets every day around the world. It's also up 71% from the BIS's last survey in 2004, the largest jump in volume since the institution began conducting its benchmark survey in 1989.

What Can and Cannot Be Predicted (and Thoughts On Telling the Difference)

I just ran across this post ("Debating the Viability of Terrorism-Prediction Markets") over at the 'Footnoted' blog at the Chronicle of Higher Education discussing the risks of terrorism how best to assess them and whether they can be known at all. (If you've grown weary hearing about war, politics and terrorism, just skip to my non-terrorist business conclusions in the last paragraph.

In a recent interview [PDF] with the Federal Reserve Bank of Richmond, W. Kip Viscusi is asked about the public-policy response to the threat of terrorism and whether we are weighing the costs and benefits in a generally rational way. "The reason this is tricky is we don’t have very good numbers on what these risks are," Viscusi says. "The estimates of the probability of a terrorist attack or the number of people who are going to die in the coming year are all over the map. So if you can’t assess the likelihood of a terrorist attack or how deadly it is going to be, it is really hard to say how much you should spend to try to prevent it." [emphasis added]

Vscusi is characterized later in the article (and not without reason) as "one of the foremost academic experts on risk". His comment is in response to a persistent school of thought that claims, as Bruce Schneier does, that from an actuarial perspective, "terrorism doesn't happen".

Without the qualifier, the idea sounds terribly cold. With all reverence for the families of those who have lost loved ones in terrorist incidents however, it's not. Businesses and governments--not to mention non-profit institutions, individuals and families--all need to assess risks as rationally as possible and take measures to hedge them. Like it or not, the value of a human life can be defined--at least partially--in dollar terms. Anyone who's ever taken a course in economics or statistics (or balanced a checkbook for that matter) has figured out that infinite spending to protect against risk is infinitely foolish and that infinite spending on risk 'A' (and no spending at all on risks 'B', 'C', and 'D') is only a variant on the same wrong-headed assumption.

That's not what's really at issue. What is at issue is the degree to which the political leanings of some lead them to believe that we can know how much to spend combating terrorism and that the 'right' number is obviously much less than we are spending today (e.g., tactical domestic measures, overhead for business, strategic overseas measures, etc.) To which my response is: really? Show me your pre-9-11 white paper predicting the order-of-magnitude sea-change that occurred in that 'industry' on 9-11.

Bryan Caplan is one of Viscussi's critics. The CHE post notes:

"I am frankly puzzled," Caplan writes at EconLog. Citing the work of John Mueller,we have a long experience with terrorism, which has "shown it to be an extremely small problem in the broad scheme of things. How much longer does Viscusi want to wait before he'll conclude that the risk is very low?"

Unfortunately, long experience in and of itself is not sufficient for prediction, even at a macro level. Hold that thought for a quick diversion.

Here's where it gets weird. In criticizing Viscussi, Caplan, an econ prof at famously free-market GMU, ends up in league with Schneier, who as far as I can tell, tends towards the opposite end of the political spectrum. Both conclude, for entirely different reasons, that the future threat of terrorism can be known and condensed to a dollar figure and that rational budgets (both public and private) can be set accordingly. Oh that it were so.

Caplan, in particular "favors the establishment of a prediction market to help assess the likelihood of a terrorist attack". That's something I can conditionally applaud. If the results are used to "help assess", then we're fine. The possibility that a prediction market might help draw in and roll up marginal, highly distributed, even intuitive information that can supplement traditional (and sadly inadequate) intelligence-gathering mechanisms is certainly a good thing.  I've been a huge fan of prediction markets for years. They should be used for more things than they are today. To my delight, more and more are catching on.

But...

...as longtime readers know, I've also concluded that there are some (and arguably many) problems for which prediction markets are not only silly but grossly misleading. I don't have the space to review them all of them here. Thinking that they can precisely predict and quantify particular terrorist threats or even the threat of terrorism generally is a notion that falls into that category. Almost by definition, a successful terrorist venture is compartmentalized, secret and surprising.

In short, Caplan, Schneier and others appear to have an ideologically-induced blind spot that leads them to declare certainty where it does not exist. Let me explain.

Schneier and Caplan draw their essential argument from a backward-looking, actuary-style catalog of terrorist incidents. This many people died. This much property was lost. Productivity was reduced by this much for this long. Etc. Etc. It all sounds very rational. If we had reason to believe that terrorism were a natural, forecastable, perhaps even cyclical phenomenon, that approach would be absolutely correct. We don't.

The main problem, as I've noted before, is that:

The  unpredictability of terrorism renders any backwards-looking, purely quantitative, actuarial mode of analysis inappropriate and ineffective. That is, future deaths due to terrorism are something that neither Schneier nor anyone else can possibly predict with any degree of confidence.

Until 2001, the biggest single terrorist incident had caused around 300 deaths. Then in the space of a few hours, that number went up by an order of magnitude. There was no consensus (or even a significant plurality) of expert opinion predicting that that would happen - much less when, where and how. [emphasis in original]

And that's the problem.

Sudden, step-function, order-of-magnitude change, precipitated by a small group that has every reason to keep its plans secret is inherently unpredictable. It can only be imagined. If that sounds familiar outside of the terrorist context, it should. Businesses face this kind of challenge all the time; it is the very nature of business, in fact: snowboards vs. skis; digital photography vs. wet-process; PCs vs. mainframes; VoIP vs. legacy telephony; biotech vs. big pharma. The list is very long. It's harder to name an industry that hasn't been touched by this kind of change at some point (often precipitating the re-invention and re-definition of the former "industry") than it is to come up with a long list of examples of industries that have be altered in this way. Bottom line: it's important to differentiate between problems that lend themselves to forecasting and those that can only be dealt with via imaginative scenarios.

Thinker or Tinker - In Pursuit of Practical Strategy

My Tuesday post ('sdrawkcaB gniknihT...', aka 'Thinking Backwards') drew this response from Dave Snowden over at Cognitive Edge. First, mea culpa for a poor choice of words in referring to author and colleague Patti Anklam whom I hold in the highest regard. On to Dave's stuff. He writes (I'm excerpting what I take to be the espresso-essence here):

1) Future Backwards [an approach he developed] is not an alternative to Scenario Planning [generic], [but] is one of a body of techniques designed to handle the issue of managing uncertain complex environments...

2) Threats and opportunities in a complex domain are unlikely to be anticipated...

3) Discussion in groups, no matter how done, constrains possibilities to our imagined present and perceivable futures. That constraint prevents us detecting weak signals or sensing different potentialities.

I'm in near-total agreement on each point:

• narratives and scenarios are complementary,
• one can never adequately anticipate the unprecedented (resilience is a far better stance) and
• groups (as well as individuals) can be remarkably sheep-like.

Here's where I suspect Dave and I are probably seeing/touching different parts of the elephant (not the unspeakable one in the living room, but rather the one in the Indian parable):

Workshop processes, especially if they are reductionist - modular piece parts [my term] implies this - will always tend the accept the dominant culture of the group, or that of the facilitation team. The field of strategy needs research and planning processes which minimise the footprint of the strategist; they also require minimising the influence of the facilitator. Focusing on participation and inclusiveness runs a very high danger of group think and consensus - exactly the opposite of what is needed under conditions of uncertainty. The role of the scenario planning consultant is not to make people comfortable, the opposite in fact. There is no point to going to great lengths to prevent the dominant Mr Smith to use Mapping Strategy's example, if the facilitator then orchestrates the assembly. Now I assume that s/he has a way of avoiding this, although I am not sure it can ever fully be escaped. In many of the scenario planning processes I have observed the facilitator in effect strongly influences the group direction.

There's more here than I can easily address in one post. One important point Dave raises is the role of the strategist/facilitator. The classic 'pure facilitator' model is much like the classic journalistic, psychiatric or pastoral ethic: one must enter an almost Zen-like zone, dispensing with (or at least setting aside) biases and assumptions in order to draw out the best input from others. Nice ideal. I buy it--to a point. Only problem: we're all human. Better to clarify, understand and declare one's biases--and let others do the same. Call it the ethic of the blogosphere: an incredibly wide range of authentic and thoroughly biased voices get heard. Nobody can say that he or she is above it all.

I'll hit a few other points via a rough analogy aimed at teasing apart the differences between narrative and scenario techniques. (I'll stick with generic terms for each method for now; other posts outline how we each believe our versions are special.) Warning: this analogy deliberately exaggerates the differences for clarity.

Developing strategy is akin to crafting a three-dimensional object out of wood. I'm consciously avoiding terms such as 'sculpture', 'model', 'artwork', 'contraption' or 'prototype' (much less 'house' or 'book') since each carries with it a set of constraints and requirements that beg a more limited set of methods. There's a time and a place for 'blank-sheet' brainstorming and sense-making out of virtually nothing (analogized at left in the elegant carved-wood 'thinker' sculpture) and a different set of conditions under which discrete representational languages are better for groups to compare, contrast and converse about a more limited set of possibilities (analogized at right in Tinker Toys).

If one's starting material is a completely raw block of wood (i.e., the cognitive vastness of wide-open  assumption-free strategy 'space') and modularity, replicability and speed are lesser priorities, then the highly refined figure at left is clearly the most useful way to think about developing strategy--and the one I'd want on my mantelpiece. Not all strategy problems are that unbounded.

If on the other hand, some basic assumptions can be reasonably made (and others dispensed with) via some front-end interviews, research, listening, narrative, analogizing and sense-making, then a rapid prototyping approach to future possibilities using modular piece-parts can enable a far greater range of strategic innovation.

This is counter-intuitive.

Other analogies may be helpful. In industry after industry (automobiles, eyeglasses, computers, software, housing, music, etc...)--and professional services are no exception: e.g., legal advice, accounting, software coding and even strategy consulting--it was once thought that every creation needed to be a 'one-off', developed from scratch to unique specifications. The results could often be near-perfection, but at a high cost. Re-configuration and plug-and-play innovation were virtually impossible.

In each case, modularization, interoperability and standardization led to a flowering of creative possibility in combining those parts. Counter-intuitive. The carved figure seems more special somehow--and it is... in one dimension (call it artistic elegance). Only a few elite folk can create such things. They cannot do it quickly. Less overall variation. Less overall innovation. Less ability to involve the people who have to live and breathe and implement the strategy.

Again, I exaggerate to tease out the differences, not to slam Dave's approach which I do not claim to fully understand. Each has its place.

The broad goals of strategy in virtually any case and with any method are to end up with something differentiable (distinct from competitors, innovative in customers' eyes and if possible not easily copied), explanatory (symbolically representing something far more complex while helping people to comprehend it), resilient (i.e., adaptable to new information, insights and changing conditions) and easily shared (i.e., enabling good ideas to be translated into common and widespread 'gut' understanding, buy-in and action)--not necessarily in that order. It strikes me that narrative methods may do better on the former two points; scenarios on the latter.

Narrative work is not all elitist fine art; scenarios are hardly flimsy children's toys (nor is the opposite true)... and all analogies and superlatives are dangerous.  ;-)

What modular, interactive scenarios do (and paleo-scenarios do far less well) is to define discrete if somewhat arbitrary points in the vastness of strategy 'space' in order to facilitate dialogue and easy recombination. Managers and executives can converse and 'play' far more easily and efficiently with discrete events ("More iPods than TVs: 2010") than they can with abstract concepts. This is the essence of resilient strategy or 'strategy-on-the-fly' (it's always evolving, never 'done'). People can express opinions about discrete evets in terms that more clearly surface differences of opinion and facilitate cognitive map-making (e.g., 'likely'/'unlikely', 'necessary'/'irrelevant', etc.)

When we do scenarios, the modular piece-parts form the basis for a common strategy language that helps groups to converse and make sense of an otherwise unmanageable set of possibilities. Narrative (in its loosest sense--i.e., people talking in stories--and not to be confused with Dave's method in particular) is a great front-end to scenarios, helping to expand and add color, definition and boundaries to a pioneerable space that's otherwise just a raw block of wood.

sdrawkcaB gniknihT - Mind Game or Creative Lever?

Related to my post yesterday about simulated hindsight, this Times of London article notes the value not only of working from a given end-point, but of actually 'telling' a story in reverse.

E.g., the firemen drove away, they packed up their gear, they put out the blaze, they broke down the door, they arrived on the scene, the mother called 911, the boy fled from his room, he screamed, his hair caught fire, he was playing with matches, he stole some matches from his father's desk drawer.

It's not easy to follow--or tell. That's the point in some contexts. Police have found it useful in attempting to catch a suspect in a lie.

Traditional police interview methods were used in the study, and in those that employed the reverse order tactics – described as “cognitive load interviews” – the interviewer asked the suspects to recall a series of events from the most recent backwards.

Officers were less likely to detect the liars when traditional methods were used in the interviews but were more likely to detect lies when the subjects were asked them to recall events in a reverse order.

It's analogous to something good proofreaders know well: spelling errors are easier to catch when one reads backwards. I was tipped off to the Times piece by innovative UK-based consultant Dave Snowden, over at Cognitive Edge. (H/T: Bob Weber) Snowden favors the backwards-telling approach because:

...by getting people to construct history in reverse... they explored more possibilities and were more open to novel discovery

That's all to the good. Snowden seems to position narrative as an improvement on scenario planning however. That's a bit hasty and sweeping.

...if you have strong opinions about what should happen, then it is easy to influence the evolution of a scenario that will support your proposed actions. Its also easy to describe how the past led to the present in such a way as to vindicate your view of history... I drew the ideas [for narrative] from various readings in the cognitive sciences which indicated that reverse time flow was harder, and disrupted what would otherwise be entrained processes. [link and emphasis added]

It's true that traditional (non-modular, non-interactive), 1970's/'Shell-style' scenario planning is particularly vulnerable to influence by strong personalities with political motives and rigid views. Any process can be steered by someone in power uncomfortable with open-ended 'novel discovery' who makes it clear to subordinates and unscrupulous consultants that there's a "right" and a "wrong" answer to the process s/he is paying for. A centralized process is 'breakable' with concerted, centralized influence. Practically every client we work with has taken us aside at some point to warn us of this in one form or another:

Before we get going, you need to know about Mr. Smith, the head of our XYZ division. If he ever gets the floor, he will kill this. Everyone will be forced to go along with his ideas. He's a smart and he'll twist this. He's a bully and a blowhard and he's got the influence to bend people to his will and make this come out his way. Watch out.

And every time, Mr. Smith (and it's usually a 'Mr.') finds himself--in our meeting--in a situation where his personal air-time is far more constrained than he's used to and his contributions are on par with everyone else. Looking for an opportunity to drive the outcome, he instead finds himself sorting through modular scenario piece-parts, unable to know in advance how the collective intelligence of 20-40 people will render them all into an interlocking scenario map.

When scenarios are built in a more distributed fashion from modular piece-parts in the context of a highly interactive, participatory, fast-moving, tightly choreographed session however, any one person's ability to steer becomes minimal. The process is democratizing, taking advantage of emergent 'wisdom of crowds' effects that nobody can see, much less steer until the end when the whole picture is developed. Some hierarchical, top-down cultures (both corporate and national) don't like our process for precisely that reason.

Story-telling is only part of it and teams will often independently elect to build a scenario story in reverse--quite possibly for the reasons Snowden suggests. His process--to my knowledge--takes advantage of similar dynamics. My contrast here is not with narrative but with traditional ('paleo-') scenarios--the kind most people are familiar with.

I've made the analogy before that traditional scenario methods are to mainframe computers as modular, interactive scenarios are to PC networks. It's not perfect, but it holds up for 'hacking' and influence as well: a network of independent actors tasked with constructing a scenario 'map' is far more resilient than a monolithic (e.g., top-down) process.

Markets make for an even better analogy (one of the reasons I continue to link them with modular scenarios). Broadly-based, highly liquid markets (including PMs) are less subject to influence than monopolies/oligopolies (one/few sellers) and monopsonies/oligopsonies (one/few buyers)  .

I don't know Snowden's process intimately, but I've been through some interesting demonstrations by one of his acolytes professional colleagues, Patti Anklam, whom I hold in high regard. Her main website can be found here. [italicized stuff added] Fascinating stuff. Powerful. And different from scenarios. Hardly a substitute for them (or vice versa!). I see two major distinctions:

1. Reverse-order storytelling, because of its "high cognitive load" (my brain hurts!), favors those with greater ability to bear such a load and/or greater experience using the technique. In other words: the same executives liable to steer any process, as well as (sometimes) the consultants. Requiring such an approach may to put a damper on inclusiveness and equal participation. Great for police questioning and trying to trip up a suspect; not necessarily so great when trying to build a team and gain their enthisiastic buy-in to a common vision.
2. The novelty-seeking open-endedness that narratives tend to produce can be just the opposite of what some organizations need. When the problem is completely unbounded (little data; near-infinite possibility; minimal understanding), scenarios are of little use and narrative may be most appropriate. When the problem is more about finding distinct points of divergence and convergence between several "schools of thought" (some data, numerous but not infinite possibilities, modest understanding) then scenarios tend to be more appropriate.