Influencing Change in a Complex Adaptive System

In a previous post, we talked about organizations as Complex Adaptive Systems and how the inherent flexibility of people allows teams to manage through immediate disruption.  This model does not only apply during disruption but offers a nice roadmap for focusing attention in a way that will lead to better outcomes by tapping into the talent of our teams as skilled humans.

In many business functions, and especially in technical support and IT Service Management environments, we have been trained to control all aspects of how people get the job done.  Measures and processes abound, with the desire to drive results.  Projects or efforts add more processes and measures without stepping back and looking at how these new demands interact with the others already in place.  Does the new initiative fit the overall outcome the organization is striving to achieve, or are they intended to fix a point issue without understanding the root cause?  Disconnects happen between the people trying to do the job and the controls, tools, and metrics that drive management’s impression of the way tasks are getting accomplished.  By adding more and more controls and measures we start driving many unintended consequences. And in the end, people find the fastest, most efficient way to get work done – often in spite of the processes and measures put on top of them in an attempt to govern behavior.

Approaching our teams as Complex Adaptive Systems is one way to rethink how we manage a business function. 

Complex Adaptive System: A system in which many independent elements or agents interact, leading to emergent outcomes that are often difficult or impossible to predict simply by looking at the individual interactions.

These three distinct components provide a way for us to look at our organizations through an adaptive lens: the emergent patterns, the independent agents, and influential interactions.   

Emergent Patterns:  Documenting the desired dominant patterns and the ones that should be eradicated or reduced is a first step in understanding the system.  The top-level purpose, mission, and strategic objectives of a company and organization will help determine the patterns (outcomes) that we would like to strengthen or achieve.  Articulating what we collectively are trying to achieve is the most important part of tapping into the abilities of an organization.  With a sound footing in the desired state, it becomes easier to document observations of patterns that are desired and ones that are not.  

Independent Agents:  In this case, agents are any of the players across an interaction including people, tools, and/or processes.  Identifying and naming the major agents helps us create a map of our ecosystem.  It is also critical to be honest about how much is known (and how much control we have) about each independent agent.  As an example, if one of the “agents” is another business function, how much do we know about the purpose, measures, goals, pain points, and needs of that other function?  It is very important to remember that our customers count as independent agents!

Influential Interaction Points:  How different agents interact is what leads to emergent patterns.  A key part of influencing an adaptive system is identifying major interaction points that are beneficial, versus those that are a liability to the desired outcomes.  We cannot possibly map every interaction, so determining which interactions are the most important or influential is necessary. 

As I started to think about an example to illustrate this, the KCS refrain “Don’t put goals on activities” kept ringing in my head. 

Here’s a scenario we hear about all the time:

  • Goal:  Increase customer access to the knowledge within our organization by increasing the number of knowledge articles available on our self-help portal.
  • Solution:  Incentivize and goal our team members to create 10 knowledge articles a month.
  • Assumption: By setting this goal and incentive, people will take the time to create knowledge articles that can be re-used both internally and externally by our customers to better resolve issues faster – a win for everyone!
  • Result:  We have created hundreds of knowledge articles: success!  But wait, these articles are not findable, not in the customer’s context, not actionable, don’t answer questions actually asked by customers, and as a result, it is now more difficult for customers to find answers through our portal.

One simple change (a goal on the number of articles created) has had a profound impact on our customers and employees, and now we need to spend added time and money to correct the error of our ways.  

This is not to say that we do not need processes, measures, or goals.  We do; they are a necessary part of any organization. But taking a more holistic approach to understanding our entire system, and focusing on how we can better create guardrails instead of overly-prescriptive workflows has profound impacts on unlocking the potential of our team.   

If we re-imagine this scenario using an adaptive approach:

  • Desired Outcome:  Increase customer access to the knowledge within our organization (Note the goal does not include ‘how’ as it does in the first scenario) 
  • Emergent Patterns: 
    • Patterns we might look for include an increase in the number of customer views of knowledge on our portal, an increase in the number of knowledge articles available through the portal, and a decrease in customer searches that return no results.  
    • Patterns we might not want to see include an increase in the number of searches performed during a customer’s session, a decrease in the amount of time per session, and/or an increase of cases opened from the portal after a knowledge search
  • Independent Agents: 
    • Knowledge Workers: create, maintain, and publish knowledge. Do knowledge workers understand the why behind the process? 
    • Knowledge Tool / CRM Tool: Are there things about the tool(s) that are interrupting the flow or presentation of knowledge? 
    • Self-Service Portal: Is the portal experience conducive to delivering and finding knowledge?
    • Customers: Do customers know where to go to find knowledge? Is it accessible and in their context?
  • Influential Interaction Points:  There are thousands of variations that take place in the interactions in this simplified example. Instead of trying to impose activities (by putting goals or incentives on specific tasks), find the interaction points that have broader influence:
    • Ensure publisher training is widely available and celebrate those who are publishing to the portal
    • Work to remove pain points from both the knowledge tool and the self-help portal and communicate those changes
    • Communicate the outcome we are aiming for and how each independent agent has an impact on the outcome – this might mean doing some customer training as well
    • Communicate and celebrate desired outcomes as they emerge!

From a big-picture standpoint, using this model as a way to influence our organizational networks can help us build resilient systems and avoid the unintended consequences of command and control thinking.  


Leave a Reply

Your email address will not be published. Required fields are marked *