Organisational psychology
Team cognitive load: reduce avoidable work
Reduce avoidable cognitive load in digital teams: clarify priorities, improve handovers and cut repeated searches without oversimplifying the work.
Before your team can solve the problem, it has to find the current brief, reconstruct a decision from chat and check whether the deadline changed. The work is demanding, but some of the demand comes from remembering how the work is organised.
That is the useful entry point for discussing cognitive load in teams. You cannot remove the thinking required by a difficult task. You can examine the extra effort created by fragmented information, ambiguous priorities and repeated interruptions.
The aim is not to make every job effortless. It is to reserve attention for judgement, learning and delivery instead of using it to reconstruct the environment. Start with one workflow and ask what people need to hold in mind to move it forward.
Use the research carefully
John Sweller’s cognitive load research examined learning and problem solving. Its distinction between task demands and the way information is presented offers a useful lens.
A workplace team is not a classroom experiment. You should not translate the research into a precise productivity percentage or claim that every meeting exceeds a universal brain limit.
Likewise, Mark, Gudith and Klocke’s interruption study showed that faster task completion under interruption can coexist with higher experienced effort and stress. Speed alone is an incomplete measure.
The recommendations here are practical work-design proposals. Use them to formulate changes you can observe in your own workflow, rather than presenting the research as proof that your preferred tool or schedule will work.
Separate necessary complexity from avoidable reconstruction
A complex task may involve uncertainty, several variables and specialist judgement. These demands belong to the work. A launch decision cannot always be reduced to a short checklist without losing what matters.
Avoidable reconstruction is different. People search for the current requirement, remember who agreed an exception or repeat a question because the previous answer is buried in a thread.
Ask someone to walk through a recent task. Note when they need domain knowledge and when they need to locate or reconcile information. Do not assume every long task is inefficient.
The distinction changes the remedy. Necessary complexity may call for a worked example, collaboration or more time. Fragmented information may call for one current record. More training will not repair a missing decision owner, and fewer meetings will not fix incomplete requirements by themselves.
Give each active task a current record
Create a place that holds the task’s purpose, owner, current scope, next decision and relevant source material. It can be an existing project tool or a simple document. The value comes from the agreement to maintain it.
Link to supporting discussion without requiring someone to read every message to understand the current state. When a decision changes, update the record and name the new consequence.
State who maintains it and when. If nobody owns the update, you have created another document that will eventually conflict with the others.
Use explicit status language. “Blocked: awaiting the approved data source from the data owner” is more useful than “in progress” when the person cannot proceed. The record should make the next action understandable to someone who was not present at the last meeting.
Make priorities a decision, not a list
A list of important projects is not a priority order. If everything is urgent, each person has to invent a rule for allocating effort and then defend it.
Name the work that takes precedence and who can change that order. Explain the consequence: which item will wait when the urgent task arrives? If the organisation needs all of it simultaneously, expose the capacity decision.
Limit active work according to the actual workflow, rather than importing a universal number of tasks per person. A support role and a research role may require different arrangements.
Use the motivation and autonomy guide when the team is asked to take initiative without permission to choose between incompatible demands. Autonomy needs a usable priority structure.
Design interruptions around the work that needs them
Some interruptions prevent harm or unblock an important dependency. Others arise because nobody knows where to put a question.
Define an urgent route for situations that genuinely need immediate attention. State what qualifies and who covers it. Let ordinary questions use a channel or review window with a reasonable response expectation.
Protect focused work without pretending that all collaboration can be batched. A person handling urgent customer issues may need a different schedule from someone analysing a complex dataset.
Make the arrangement visible to colleagues outside the team. Otherwise a protected period becomes a local preference that other functions accidentally override.
Review whether the rule creates delayed assistance or shifts all interruptions to one person. Reducing one colleague’s switching cost by overloading another is not an improvement in the whole workflow.
Improve handovers by showing the next decision
A handover should tell the receiving person what they need to do, which assumptions remain and where the current evidence lives. Sending a folder of materials transfers information without necessarily transferring a task.
Ask the receiver to describe their next action. If they cannot do so, the handover is not complete. Add missing context while the sender is still available.
Include an example of the expected output when the task is unfamiliar. Distinguish an example from a rigid template: the recipient still needs to know which parts can change.
The organisational psychology guide helps connect these handovers with decision authority. Unclear responsibility often creates repeated information work.
A hypothetical reporting workflow
Imagine a team producing a weekly portfolio update. This is a hypothetical example. The source sheet lives in one folder, project decisions are in chat and the manager requests revisions by email.
The team maps one update. Much of the effort goes into reconciling versions and asking whether a deadline change is approved. The underlying portfolio analysis still requires expertise; that is not the part they try to eliminate.
They introduce one current project record with a named maintainer. Decisions link back to their source discussion. Each project owner supplies a short exception note before the reporting review. Urgent changes have a separate route.
The team compares repeated clarification requests, corrections caused by old information and total preparation effort. It also checks whether maintaining the new record creates additional duplication.
The outcome is deliberately open. If the record merely adds another reporting obligation, the design needs to change. The example illustrates how to isolate avoidable reconstruction without claiming that the analytical work will become simple.
Use this cognitive-load audit
Choose a workflow, not an employee. Complete the audit through observation and conversation, with people aware of its purpose.
Swipe or use the arrow keys to see every column.
| Question | Look for | Possible change |
|---|---|---|
| What must be remembered? | Hidden rules, dates and approval conditions | Put them in the current task record |
| What must be searched? | Sources spread across channels | Link the approved source at the point of work |
| What must be reconciled? | Conflicting versions or priorities | Name the owner of the current version |
| What repeatedly interrupts? | Ordinary questions treated as urgent | Agree response routes and expectations |
| What is genuinely difficult? | New concepts or specialist judgement | Add examples, practice or collaboration |
| Where does work return? | Handover gaps and misunderstood criteria | Confirm the receiver’s next action |
For each change, specify the owner and the review point. Do not launch six improvements at once unless you can understand their combined effect and resource the work. A smaller change often produces a clearer operational lesson.
Watch what AI adds to the workflow
An AI tool can shorten the creation of a draft while adding source checking, correction and approval. Count the whole sequence.
Make the reviewer’s responsibility explicit. A vague requirement to “check the AI” may create more uncertainty than it resolves. Define which elements need verification, which sources are permitted and what happens when the output fails.
Avoid placing a generated summary beside an authoritative record without distinguishing their roles. If people cannot tell which one governs the task, you may have introduced a new version problem.
The guide to resistance to AI change explains why reluctance can contain valid observations about this additional effort. A colleague who questions a supposedly faster process may be accounting for the work after generation.
Measure the arrangement without monitoring every person
Useful evidence includes rework caused by outdated information, clarification loops, waiting for decisions and the time needed for a complete task. Voluntary feedback can add information about effort and predictability.
Avoid installing invasive activity monitoring to infer cognitive load. Screen time and message volume do not reliably describe the quality or difficulty of thinking. A person reading quietly may be doing essential work.
Record changes in workload and staffing when comparing periods. A lighter week can look like a process improvement; a difficult release can hide a useful change.
Do not make “no questions” a target. Questions may prevent a serious error. The psychological safety guide helps keep clarification available while improving the environment.
Start with the information people keep reconstructing
Ask the team which question has been answered more than once this week. Find the missing record or unclear decision behind it. Make one source current, agree who maintains it and review whether the repeat work decreases.
The same information-design principle appears in consumer psychology: an observed difficulty suggests a question about the environment, not immediate evidence of a personal limitation.
If an AI or digital programme is spending too much time finding context, tell me which handover requires the most reconstruction. We can redesign that step so the team can spend more attention on the work itself.
Apply this to your project
Tell me the decision you are facing, the constraints and what you have already tried.
Keep exploring
Motivation at work: give autonomy a clear shape
Understand motivation and autonomy at work: clarify choices, support competence and design responsibilities people can exercise in practice.
Organisational culture: look at everyday behaviour
Understand organisational culture through decisions, rewards and everyday responses, then turn stated values into working practices you can review.
Organisational psychology: a practical team guide
A practical guide to organisational psychology for digital teams: understand friction, clarify decisions and improve how people work together.