Introduction

A feedback loop exists if a ToC element that in time comes after another ToC element, feeds back to the first element, thereby forming a circuit or loop. For example, an output may produce an outcome or change the context, while the outcome or context may again feed back into the input that produces the output.

For example, an organisation may conduct training that contributes to participants’ willingness to experiment with new techniques. A year later the organisation engages participants as trainers in a new training program. This way lessons-learned by participants during that year feed back into the training program, thereby further contributing to innovation.

Why Important?

Pathways of change can include feedback loops. In fact, feedback loops are essential to a good ToC. Without feedback loops a ToC is reduced to a linear process, which is not likely to be a proper representation of reality as social systems contain dynamic and circular processes.

Feedback loops can accelerate the change process and provide a source of competitive advantage and growing success. This occurs when they produce reinforcing effects within a positive change process.

A reinforcing effect occurs when an initial change is (automatically) reinvested to further that change in the future: element A leads to element B, which (automatically) reinforces element A, which again further increases element B, and so on. Such thus accelerates the change process.

Feedback loops can also have a balancing effect. Such an effect stabilizes the change process by countering a change with a push in the opposite direction.

A balancing effect occurs when an increase in element A increases element B, which reduces element A, which again reduces element B, creating a balancing level between the two ToC elements.

Word of Caution

We advise against using so many feedback loops that the logic becomes meaningless. Everything is related to everything else, but this can make it very difficult to understand. Nevertheless, some feedback loops can be critical to the success of a program and should be included.

Illustration

The illustration below is loosely based on a ToC for a training program for survivors of domestic violence, developed as ‘Project Superwoman’ by ActKnowledge and the Aspen Institute.

Several feedback loops can be identified:

Increased impact (1a) and adoption (1b) lead to a better reputation, visibility and word of mouth, thereby again furthering enrolment in the program.
Another feedback loop shows a learning process. As more training sessions are provided, the organisation learns about how to optimally give form to such training. To ensure learning, may require its own actions such as strategies to incorporate trainee feedback in the development of the training materials.
It also illustrates a balancing effect, namely in the form of learning potential that sets limits to the extent of the learning effect.

Sources of Feedback Loops

Potentially important sources of feedback loops include:

Learning Effects: as activities and output grow, the organisation continues to learn and grow in effectiveness and efficiency.
Scale Effects: (production) cost per unit goes down as output increases.
Network Effects: as the user base grows in size, more compatible propositions become available thereby adding to the original proposition’s value.
Social interaction Effects: stakeholders value a proposition higher as more users adopt the proposition, for example because of increased brand recognition and social signalling.

Questions to Ask

The following questions can help to identify feedback loops:

How are our resources and capacities further built, improved or limited as we grow in success?
Why would it become easier or more difficult to execute our ToC as we become more successful?
Do we build outcomes that again influence the context thereby building a context supportive of our vision of success?
How do we and other actors within our ToC’s scope, build a competitive advantage and how is such related to our ToC?
Do any of the following effects play a role over time for our ToC: learning effects, scale effects, network effects or social interaction effects?