An analyst working on a competitive landscape report has, open at any given moment: the draft document, a spreadsheet of company metrics, three or four PDF reports from industry sources, eight browser tabs with recent coverage, a notes file from earlier in the project, and possibly a communication thread with a colleague. This is not unusual. It is the ordinary environment for knowledge work that draws on multiple sources.
Every time the analyst switches between these contexts, something specific happens in their cognition. The mental model they were holding about the draft's current argument, or the specific data point they were tracing across sources, has to be rebuilt. Partially. The new context displaces part of what was being held. The cost is not the seconds of the switch itself. It is the degraded quality of attention that follows it.
What Context Switching Actually Costs
The cognitive research on task-switching has been fairly consistent for decades: switching between tasks of any meaningful complexity carries a re-entry cost. The person returning to a task after switching away takes time to reconstruct the mental state they held before the interruption. The more complex the task, the longer the re-entry takes, and the more incomplete it tends to be.
For analysts, the relevant complexity is not just the individual task (reading a PDF, updating a spreadsheet) but the higher-order task: building and maintaining an argument that draws on content across all of those sources simultaneously. Each time the analyst switches from the draft to a source, or from one source to another, they partially exit the argument-construction task and partially enter a source-engagement task. Each time they return to the draft, they re-enter an argument-construction task that is a slightly degraded version of where they left it.
The accumulated effect across a four-hour working session with frequent switching is a persistent state of partial attention, in which no single context is ever fully inhabited. This is the environment in which the connections that make good analysis get missed: the implication that would have been obvious if both sources had been held in mind at once, but was not, because attention was divided.
The Source-Tab Problem
The browser tab as a unit of work deserves specific attention, because it is where the context-switching problem is most visible and where it compounds most quickly.
An analyst keeping 12 browser tabs open is maintaining 12 potential context switches. Each tab represents a source, a search result, a piece of coverage, a reference document, or a tool. Some of those tabs are genuinely active; others are parked on the basis that they might be needed and are too costly to find again if closed. The result is a working environment that presents constant implicit demands on attention, even for the tabs that are not currently active.
The deeper problem is that tabs are organized spatially (position in the tab bar) rather than semantically (relevance to the current argument). When an analyst needs a specific piece of information from a source, they must either remember approximately which tab contains it, or scan through the tabs to find it. Neither of these is a zero-cost operation when the analyst is also trying to maintain the thread of an argument they are writing.
The Transition Cost Is in Both Directions
There is a useful asymmetry in how the context-switching cost distributes. The cost of switching from the draft to a source is relatively low: the analyst is exiting a high-complexity context (argument construction) to enter a lower-complexity one (source retrieval). The cost of switching back from the source to the draft is higher: re-entering argument construction from a source context requires reconstructing the state that was interrupted.
This asymmetry has a practical implication: the number of times an analyst must leave the draft to find something is directly proportional to the amount of time they spend in partially-reconstructed argument-construction states. An analyst who must switch to a source context five times during a writing session faces five re-entry costs. An analyst who can retrieve five pieces of information without switching context faces none.
This is not about discipline or focus. It is about whether the workflow requires the analyst to leave the draft at all to find what they need. If all the relevant sources and their extracted content are accessible without exiting the current context, the switching cost does not occur. If they are not, it does.
Multi-Project Switching Compounds the Problem
Many analysts work on more than one project simultaneously. A policy analyst at a growing think tank might be actively writing one brief while also doing early-stage reading for a second and following up on revisions for a third. This is not unusual, and it creates a context-switching problem that operates at a higher level than the within-project switching described above.
Each project has its own working context: its own sources, its own argument structure, its own vocabulary and frame. Switching between projects requires not just a context switch between tasks but between entire mental models. The analyst moving from one brief to another does not simply shift attention; they re-instantiate a different set of background knowledge and associations.
The cognitive cost of this higher-level switching is substantially larger than within-project switching. It is also less visible, because the analyst may not experience it as switching at all. They are just working. But the depth of attention available for each project is reduced by the overhead of maintaining multiple working contexts simultaneously.
What Reduced Switching Buys
The claim that fewer context switches improve work quality is not difficult to accept in the abstract. The harder question is what "better" means in concrete terms for analysis work.
The most tangible benefit is not speed, though analysts working in reduced-switching environments do tend to work faster on individual documents. The more important benefit is depth: the sustained engagement with a single argument for long enough to notice implications that require holding multiple things in mind simultaneously. The insight that comes from comparing two sources directly, rather than sequentially, each already fully loaded in working memory.
We are not claiming that context switching can be eliminated from analytical work. Some switching is inherent to any work that draws on multiple sources. What can be reduced is the switching triggered by logistical failures: not being able to find something, needing to re-read something that should have been annotated, exiting a document to locate a reference that should already be available in context.
That is the category of switching that Pilgrim is designed to remove. The kind that is not intellectually necessary but operationally required by the way sources and notes are typically organized. When the relevant excerpts, connections, and annotations are already assembled in one place before the writing session begins, the analyst can stay in the argument for longer, and the quality of the analysis reflects that.