A quick review of a second article in Harvard Business Review this month on Complexity.  This one is “Learning to Live with Complexity” by Gökçe Sargut and Rita Gunther McGrath.

See my earlier post on “Embracing Complexity” from the HBR.   Before getting too far into this subject, I want to reference an excellent E-Book by Tom Koulopoulos on this subject, called “The Uncertainty Principle.”  In this downloadable PDF he points out that  uncertainty is not the enemy and that uncertainty = opportunity.  If you have not read this, go get it, it is a gem.

Also, WfMC is releasing a new book “Taming the Unpredictable” this month which expands upon Mastering the Unpredictable and also includes real world case studies.  I hope to have a detailed review of this book within the week.

My theme for the past two years has been “Unpredictability” (OMG has it really been two years!)  Unpredictability refers to the ways that a complex system can not be predicted from the initial conditions.  My favorite quote from this article is:

Practically speaking, the main difference between complicated and complex systems is that with the former, one can usually predict outcomes by knowing the starting conditions. In a complex system, the same starting conditions can produce different outcomes, depending on the interactions of the elements in the system.

Many people then assume that if a system is not predictable, that it must entirely random.  That is not the case either.  There is predictable, there is random, and there is a third option: stable, adaptive.

The [complex] system is predictable not because it produces the same results from the same starting conditions but because it has been designed to continuously adjust as its components change in relation to one another.

Then we get to the heart of what I have been saying about Adaptive Case Management: you don’t approach a complex system by modeling it!

It’s possible to understand both simple and complicated systems by identifying and modeling the relationships between the parts; the relationships can be reduced to clear, predictable interactions. It’s not possible to understand complex systems in this way, because all the elements are interacting continuously and unpredictably.

This is the message that I have been trying to get to the OMG case management modeling notation working group: model driven design does not work for complex systems.   That does not mean that it is useless: businesses are conglomerations of complicated and complex systems.  The complicated parts can and should be modeled.  The complex parts should not be modeled, and instead left for a real intelligence to work out at run time.

In a complex environment, even small decisions can have surprising effects.

This is the danger of the “just pick a process” attitude with the assumption that any documented process will be better than no process.  That is false in the face of a complex system, because picking a process implies many decisions about how to proceed, but those decisions are made outside of the context of the situation, and may have very surprising effects.  The authors identify three situations where this presents itself:

1. when events interact without anyone meaning them to
2. unintended consequences that are based on an aggregate of individual elements, not a single occurrence
3. when policies and procedures remain in place long after the reason for their creation becomes obsolete

This last reason is a strong argument against fixing the way we work today into a concrete process — because it might be just the wrong thing tomorrow.  What can we do?

Embedded in many analytic tools are two assumptions that don’t hold for complex systems. The first is that observations of phenomena are truly independent. …  The second is that it’s possible to extrapolate averages or medians to entire populations.

These are pervasive misconceptions; they hold in complicated systems, but not in complex systems.  Inability to distinguish the difference is the fatal flaw.  You can’t convert a complex system to a complicated one by analytics alone.  This quote pretty much precisely sums up how to respond:

In an unpredictable world, sometimes the best investments are those that minimize the importance of predictions.

To someone with a Newtonian mindset, this sounds like giving up, throwing in the towel, and just not even trying to do a proper job, and many traditional managers will find this unacceptable.  However this may be more palatable: instead of predicting exactly what kind of consultant we will need for the next year (and putting them on the payroll) instead build a network of consulting partners, learn what each is good at, and bring in the right ones at the right time.  You do not completely abdicate responsibility for preparing for the future, but you prepare in a way that gives you flexibility and agility in the future.  In Mastering the Unpredictable I talk about “highly reliable organizations” who do not attempt to predict the future, but instead carefully prepare a response for any possible future situation.

The article contains a lot of good advice for dealing with complex systems, and I can’t cover it all here.  These final two quotes are excellent:

Complicated systems are like machines; above all, you need to minimize friction. Complex systems are organic; you need to make sure your organization contains enough diverse thinkers to deal with the changes and variations that will inevitably occur.

business life has always featured the unpredictable, the surprising, and the unexpected.

The article also says that due to information technology, life has become more complex because things are far more interconnected than before. I speak often about how business process automation has removed the simple work, and left the hard work, making a higher percentage of people knowledge workers. Perhaps this is the same thing, or perhaps there is a new effect that I was not formerly aware of, making life for the average worker fundamentally more complex than it was 10 or 20 years ago.

All of this leads to a clear need for technology to help knowledge workers, and regular readers will know I am going to make a plug for Adaptive Case Managements at this point, but that is because it makes sense. You can not make complexity go away by making automation more powerful. What is needed instead is true intelligence (people), and systems to support people.