What is Disfluency Theory?

Disfluency Theory is a psychological and educational concept that suggests introducing deliberate challenges, difficulties, or “frictions” in a process can enhance learning, comprehension, and performance. The idea is rooted in the principle that when tasks require more cognitive effort, they promote deeper engagement and longer-lasting retention.

Don’t Automate Waste

In a world that celebrates efficiency and automation, the concept of intentionally introducing difficulty into a process may seem counterintuitive. However, there is a growing recognition that sometimes, making things slightly more challenging can lead to better results. This counterintuitive approach is known as “disfluency,” and it challenges the prevailing notion that automated processes are always superior. In this blog post, we’ll explore the idea of disfluency, emphasizing the importance of not automating away essential aspects of a task. As the saying goes, “Don’t automate waste.”

Is your data fluent or disfluent?

Disfluency is to make something more difficult in order to absorb it and then to take action. It’s the effort of turning data into power. It’s not enough just to have information. When we want to learn from information, we should force ourselves to do something with the data.

“Fluent” processing allows us to take in key information quickly—but not necessarily to retain it or even understand it in a meaningful way. The whole experience becomes meaningless, less engaging, and unsatisfying.

Conversely, we process disfluent information more carefully and deeply, and this naturally results in us understanding it better.

What does that mean? When you find a new piece of information, force yourself to engage with it, to use it in an experiment or describe it to a friend—and then you will start the connections that are at the core of learning.

An example: the daily huddle

Every morning, my organization has a huddle. Each team member has numbers they have to report to the group on a huddleboard. For me, this involves counting leads, web traffic, conversions, and other marketing data. Most of these numbers have to be chased down and calculated and takes a small amount of time. I have often wished that all this data was just automatically found and reported on a dashboard. That would be easy—much less effort, right?

But the reality is, if I wasn’t putting in the time and effort to personally know the numbers, I wouldn’t be as aware of problems, trends, or successes—and how I should be spending my time and energy.

Having me grapple with these numbers on a daily basis is called disfluency. This means that people process information differently, and that some of it is easy (fluency) and some of it requires effort (disfluency).

Should we make information more disfluent?

A lot of the data that we use day to day needs to be fluent. We need to be able to access and use it quickly, so it should be easy to digest. However, information that is easily consumed is also easily forgotten.

In almost everything we do there are a few key measures that tell us how we are doing against our goals and targets. Data such as production data, sales information, or financial projections need to move beyond abstract numbers and become more intuitive, becoming much more central in our awareness, moving from organization knowledge to personal understanding. It’s this data that needs to be deeply understood so that it can underpin the decisions we make.

How can organizations improve the comprehension, retention and impact of data?

It’s a good idea to look at the fluency of key data or information within your organization. If it’s being presented to people too easily, consider the following suggestions:

  • Asking for reports that require some small amounts of manual work to create, such as looking stuff up. An example is entering a number into a scorecard as a way if indicating that the person “touched” the data.
  • Ask people to interpret data, not just produce it. An example of this is asking the members of a huddle, tier meeting, or business review to interpret the number and what it means, such as “what is missing from this data? or what does this metric not show us?”
  • Ask data providers to add a footnote to the data. An example of this would be: “What explains the current status? Why is this happening, do we need to investigate more?
  • Draw a picture. Asking someone to make a picture from the information requires an additional layer of thinking and can instigate additional insights and value add.

Does Automating Data Capture Add Value?

Part of being a lean organization means eliminating waste. Why would we want to intentionally leave an extra step in a process that could be “automated” out?  The only Lean reason is: if that step is a value-add step. How can you decide if manually entering the numbers is value add or waste?

Manually entering data may be Value Add if:

  • The person entering the data can influence the output with their efforts.
  • The amount of time required is minimal. One number per person per day is a good rule of thumb.
  • If an audit of the data is required, and it can be performed as part of a manual cut/paste process step.
  • The data is entered by someone who can learn from the data, or may be able to suggest process improvements or recognize correlation.

Manually entering data may be WASTE if:

  • The person doing the data entry is disconnected from the work that the number represents.
  • The person entering the data is NOT able to influence the output.
  • The amount of data is very large. If one person is cutting/pasting 100s of numbers, it is probably waste.

So. Much. Data! You’ve heard the stories about how much data is available online—and how much more is added every day. And even though all that information is available to us, humankind hasn’t grown all that much smarter. This is because we don’t know which data is important and which is trivial, we’ve gone blind to almost all of it.

It’s an uphill climb from data to information to knowledge to wisdom to action—and that uphill climb is called “disfluency.”

Disfluency Examples:

Training and Onboarding

  • Example: Incorporating hands-on problem-solving tasks in training sessions instead of providing step-by-step instructions. This forces new employees to think critically and understand the “why” behind each process.
  • Benefit: Enhances retention and ensures that employees can troubleshoot independently.

Quality Control Checklists

  • Example: Requiring operators to manually verify a checklist before approving a batch rather than automating the approval entirely.
  • Benefit: Promotes careful inspection and reduces the likelihood of overlooked defects.

Material Handling

  • Example: Introducing a manual verification step in material requisition processes, such as requiring a second operator to confirm part numbers or quantities before moving items.
  • Benefit: Reduces errors in part usage or stock mismanagement.

Safety Protocols

  • Example: Requiring workers to manually confirm safety equipment settings (like lockout/tagout) instead of relying solely on automated systems.
  • Benefit: Increases awareness of safety protocols and reduces complacency.

Inventory Management

  • Example: Conducting physical inventory counts periodically, even if a digital system tracks inventory levels.
  • Benefit: Highlights discrepancies and encourages team accountability for accurate record-keeping.

Shift Changeovers

  • Example: Implementing detailed handoff procedures where outgoing employees must explain their work progress and potential issues to incoming shift members rather than relying on automated logs.
  • Benefit: Improves team communication and ensures continuity