Structural Friction Study
Quantify the exact percentage of your team's capacity consumed by coordination overhead, and identify which category costs you most.
5 minutes Take the AssessmentFour outputs from one 5-minute assessment.
How much of your team's capacity is lost to coordination rather than actual work, expressed as a concrete percentage.
A financial estimate based on your team size and typical salary ranges, translating overhead into a number your leadership will recognize.
Your breakdown across three dimensions: activation friction, knowledge friction, and decision friction, showing exactly where the drag originates.
Which of five categories costs you most: context reconstruction, decision re-litigation, manual routing, status synchronization, or knowledge loss.
The study takes approximately 5 minutes. The tradeoff pairs pit each friction type against the others, so your responses reveal which friction dominates your experience. The scenarios present two recognizable workplace situations (a stalled project and a departing colleague) and ask you to diagnose, respond, and imagine future solutions.
Includes 2 best-worst scaling blocks and 4 rapid rating items. A confidence check follows each scenario question. No sign-up required. Results are instant.
AI is being deployed to speed up knowledge work, and in many cases it is succeeding at the task level. Individual tasks are faster. But organizations are not faster. In some cases, they are slower, because AI amplifies the production of work that the organization's structure cannot absorb.
This study investigates why. We measure three types of structural friction that exist independent of AI but that determine whether AI deployment helps or hurts:
Activation Friction (from Chapter 1 of Proxy.Me). Work stalls not because the task is hard, but because starting it requires assembling information, permissions, and people. The delay is in getting to the point where work can begin.
Knowledge Friction (from Chapter 2). Expertise is locked in individuals. The organization cannot access what it collectively knows without finding the right person and hoping they are available. When that person leaves, the knowledge leaves too.
Decision Friction (from Chapter 3). Decisions are made, communicated, and then lost. The same decision gets revisited because the reasoning was never captured. New team members cannot learn why things are done a certain way because nobody recorded the "why."
These three friction types interact with AI deployment in specific, measurable ways. AI deployed on top of high activation friction just produces more work that gets stuck in the queue. AI deployed in an environment of high knowledge friction produces outputs that lack the context only experienced humans hold. AI deployed despite high decision friction accelerates choices that will be relitigated next quarter.
This study operationalizes the diagnostic framework from Chapters 1, 2, and 3 of Proxy.Me. The core argument of the book is that AI deployed without addressing structural friction accelerates the wrong things. This study tests that argument empirically by measuring both friction profiles and AI effectiveness (through a calibration item that asks whether AI tools have reduced friction in the respondent's work). The relationship between friction type, friction intensity, and AI effectiveness is the central research question.
You will receive one of 15 profiles across four categories:
Single-Friction Dominant profiles (such as The Relay Runner, The Deep Expert, and The Decision Archaeologist) identify the specific sub-type of friction that most affects your work.
Dual-Friction Patterns (such as The Coordination Catalyst and The Institutional Decoder) capture compound problems where two friction types reinforce each other.
System-Wide profiles identify whether friction is pervasive or genuinely absent.
The Paradoxes (such as The Rapid Responder and The Hidden Bottleneck Finder) reveal patterns where your experience of friction does not match what the data suggests, which is often the most valuable insight of all.
A key finding for many participants is that AI tools have improved their individual task speed without touching their dominant friction type. This disconnect explains why people feel productive but frustrated at the same time.
You will also learn the solution type you would value most: whether you would benefit most from workflow routing, knowledge management, decision documentation, or real-time visibility.
5 minutes. No sign-up. Instant results.
Take the AssessmentComplete all three assessments for your AI Position Map: