Batching Work: The Research Case for Processing Similar Tasks Together

Rubinstein, Meyer and Evans (2001) quantified task-switching costs: switching between tasks adds switch costs (time and error rates) that accumulate throughout the day. Batching similar tasks reduces the number of switches, lowers total switch costs, and preserves the warm cognitive state built during sustained work on one type of task.

2026-02-19 6 min read

The Task-Switching Cost Research

Joshua Rubinstein, David Meyer, and Jeffrey Evans published “Executive control of cognitive processes in task switching” in the Journal of Experimental Psychology: Human Perception and Performance in 2001 (27(4), 763–797). The paper used controlled laboratory experiments in which participants switched between well-defined tasks to isolate and measure the cognitive cost of the switch itself.

The researchers identified two components of switch cost: goal reconfiguration (the time required to set the mental agenda for the new task) and rule activation (the time required to bring the cognitive rules for the new task into active working memory). Both components contribute measurable additions to response time and error rates after a switch, with the effects largest when tasks are more complex and more different from each other.

Switch costs compound

Task switching imposes two-component cognitive costs (goal reconfiguration and rule activation) that add measurable time and increase error rates. Each switch in a day accumulates these costs. For complex knowledge work, the switch cost can represent a significant fraction of total productive capacity.

Rubinstein, J.S., Meyer, D.E. & Evans, J.E. (2001). Journal of Experimental Psychology: Human Perception and Performance, 27(4), 763–797.

Gloria Mark and colleagues at UC Irvine have studied interruption and recovery in naturalistic knowledge work settings. Their research has documented that workers interrupted during a primary task take time to fully re-engage with the original work after an interruption, and that the re-engagement time is often longer than the interruption itself. Importantly, Mark’s laboratory and interview research is separate from the controlled task-switching experiments; the specific figures often cited in productivity literature vary by study and should not be attributed to any single paper without care.

What Batching Reduces

Batching reduces task-switching costs by scheduling similar tasks in dedicated windows rather than distributing them throughout the day. The mechanism works through several channels:

Fewer switches per day. If email is processed twice daily rather than continuously, the total number of email-to-other-work and other-work-to-email switches drops significantly. Each eliminated switch removes an associated switch cost: added time and errors that do not occur.
Warm cognitive state. After working on a type of task for some time, the relevant mental models, vocabulary, and context are active and accessible. The first email response in a batch benefits from the warm state built by the previous responses. Starting each email independently, interspersed with other work types, rebuilds this warm state from scratch each time.
Reduced partial attention residue. Sophie Leroy’s research on “attention residue” (2009, Organizational Behavior and Human Decision Processes) found that when people move from one task to another, thoughts about the first task continue to occupy working memory, reducing the cognitive capacity available for the second task. Completing tasks within a batch before moving to a different type reduces unresolved attention residue.

Practical Batching Strategies

Communication batching. Processing email, Slack, and messages in defined windows (e.g., 9 AM, 1 PM, 5 PM) rather than continuously reduces the interruption frequency and eliminates the reactive context switching that distributed message monitoring creates. The tradeoff is response latency: batching requires setting expectations about response time norms and being explicit with collaborators about communication availability.
Meeting batching. Grouping meetings into dedicated days or windows (e.g., meetings on Tuesday and Thursday, focus work on Monday, Wednesday, Friday) creates extended uninterrupted periods for deep cognitive work. The switch cost from meeting preparation and context-switching between meeting and non-meeting work is reduced when meetings are clustered rather than distributed across all days.
Administrative batching. Expense reports, scheduling, document reviews, and other administrative tasks share a cognitive mode that is different from creative or analytical work. Grouping administrative tasks into a dedicated window (e.g., Friday afternoon) allows this mode to warm up and carry through multiple tasks rather than being reactivated repeatedly throughout the week.

Frequently Asked Questions

Does batching work for urgent or time-sensitive communications?

Batching requires explicit triage to distinguish between genuinely urgent communications and communications that merely feel urgent. Most professional emails and messages are not genuinely time-sensitive; they are processed continuously out of habit, social norm, or anxiety rather than because they require immediate response. For genuinely urgent communication (defined as: a real cost occurs if response is delayed more than X hours), batching can be combined with an urgent channel (direct phone call, specific subject line flag) that is monitored more frequently. This preserves the batching benefit for the majority of communication while maintaining a pathway for genuine urgency. The research on email processing shows that most email that feels urgent has a real response window of several hours, not minutes.

How does batching interact with reactive work environments where interruptions are frequent?

Batching is most effective when the person has some control over their interruption environment. In environments where interruptions are externally generated and cannot be deflected (open-plan offices, always-on communication cultures, high-urgency client-facing roles), the structure required for batching is harder to create. Partial batching strategies that don't require absolute protection can still reduce switch costs: processing email twice rather than twenty times per day produces some benefit even if it doesn't achieve deep focus protection. The prerequisite for more substantial batching is agreement with collaborators and managers about response time norms, which is as much a culture change as a personal productivity practice.

What's the evidence for how long it takes to fully recover from an interruption?

The specific '23 minutes to recover from an interruption' figure circulates widely in productivity literature, but it does not come from a single peer-reviewed paper. It originates from Gloria Mark's researcher interviews and discussions rather than from a formally published controlled experiment. Mark's 2005 CHI paper (with Gonzalez and Harris) reported that workers spent an average of approximately 3 minutes on a task before switching to another, not that recovery took 23 minutes. The broader research evidence from the task-switching cost literature and attention residue work supports the principle that interruptions impose meaningful re-engagement costs, but specific recovery time figures vary significantly by task type, interruption type, and individual differences. The practical implication is that interruptions are genuinely costly, even if the exact cost depends on context.