Research tests the Polaris Shift thesis against observable changes in customer behavior, interfaces, economics, organizational design, and competition. The goal is to distinguish durable structural change from technological noise.
A DEVELOPING BODY OF THOUGHT
An idea should earn its place.
Polaris Shift is being developed through a working argument, practical tools, and a planned research program. The goal is to find where the framework helps, where it overlaps existing thinking, and where it fails.
AVAILABLE TO EXPLORE
The working argument.
The first white paper brings together the framework, research context, fictional decision rehearsals, and a practical review guide. It presents a conceptual framework, not a completed empirical study.
Reader tests, practitioner conversations, and application trials should examine whether Polaris improves a real strategic discussion. These activities are planned; no findings are claimed here.
Before the next white paper: document methods, counterexamples, intellectual lineage, and the limits of the evidence.
The book should develop the argument through researched cases, competing explanations, and practical use. Its scope should follow what we learn. No publication date is announced.
Leaders find that ordinary planning questions surface the same issues just as clearly.
Readers repeatedly confuse a new measurement with a different strategic outcome.
Application trials produce interesting conversation but no useful change in the decision or next test.
Examples support adapting execution far more often than reconsidering the guiding outcome.
These are questions to investigate, not research findings. Narrowing the claim or refining the method would be a useful result.
How should evidence be presented?
Separate an observed event from its interpretation. Label fictional examples and hypotheses. Publish the context and limits of research, include counterevidence, and distinguish a participant’s report from an independently verified outcome.
The four-step method remains SIGNAL → SHIFT → STAR → SYSTEM. Evidence informs the learning loop after SYSTEM.
Polaris Shift treats evidence as a way to challenge the thesis, not decorate it. Select a domain to see the questions and observable changes the research program should track.
01
QUESTION WE ARE TRACKING
What are people beginning to delegate to AI that they previously did themselves?
Behavior matters because strategic value can migrate before an organization's internal systems or metrics notice the change.
EVIDENCE TO WATCH
Changes in discovery and decision behavior
Delegation of actions, not only information retrieval
Trust and repeat-use patterns
Journey stages that shrink or disappear
THE STANDARD
A signal becomes strategically interesting when evidence suggests a durable change in behavior, economics, interfaces, work, or competition, not merely a new technical capability.
Measurement discipline
AI evidence should be triangulated, not worshipped.
Usage, sentiment, and outcomes each reveal something different. No single stream is enough to tell us whether AI is actually creating value.
TELEMETRY
Useful for observing behavior. Weak as proof of value.
Usage, workflow activity, acceptance rates, cycle times, and other system events show what people and tools are doing. They are valuable signals, but activity can rise without quality, trust, or customer value improving.
VARIANCE LENS
The average can be true and still hide the story.
Aggregate AI impact can look positive while individual teams, workflows, or markets move in opposite directions. Research should ask who improved, who deteriorated, under what conditions, and why.
Never publish an average without interrogating the distribution underneath it.
NEGATIVEPOSITIVE
TEAM AImproves
TEAM BDeteriorates
TEAM CMixed
AVERAGELooks modestly positive
Illustrative distribution, not empirical data.
Evidence discipline
When does a signal deserve strategic weight?
Polaris Shift should not confuse novelty with structural change. A signal becomes more strategically meaningful as evidence accumulates across repeated behavior, economics, interfaces, and competitive response.
RESEARCH PRINCIPLEConfidence should rise with evidence, not excitement.
01 · EVENT
Something new happened.
A model launches, an agent gains a capability, a platform changes an interface, or a company introduces a new behavior. This is worth noticing, but it is not yet proof of a strategic shift.
EVIDENCE STANDARD
Confirm the event through direct or primary sources. Separate what was demonstrated from what was claimed.
CONFIDENCE LANGUAGE
“A signal worth watching.”
THE DISCIPLINE
Do not infer adoption from capability. Do not infer behavior from a launch announcement. Do not infer a North Star shift from one impressive demo.
Editorial provenance
Know what kind of claim you are reading.
Polaris Shift separates original concepts, external evidence, strategic interpretation, and hypotheses. The distinction matters because each deserves a different standard of proof.
01Original concept
A framework, term, or model developed as part of Polaris Shift.
Explain the logic clearly and distinguish the concept from supporting evidence.02Source-derived evidence
A factual finding, observation, or data point drawn from an external source.
Cite the source, preserve its scope, and avoid extending the claim beyond what the evidence supports.03Polaris interpretation
A strategic inference Polaris Shift draws from one or more pieces of evidence.
Show the reasoning, use confidence language, and make counterevidence visible.04Prototype or hypothesis
An idea worth exploring that has not yet earned the status of a validated conclusion.
Label it explicitly, identify what would strengthen it, and identify what would weaken it.
POLARIS STANDARD
Do not make interpretation look like evidence. Do not make hypothesis look like conclusion.
Polaris hypothesis · Strategy lens
The force that changes the forces.
AI may be strategically important not because it adds a neat sixth box to Porter's Five Forces, but because it can alter the economics and behavior inside all five at once.
ALGORITHMICLEVERAGE
COMPETITIVE RIVALRY
Competition can accelerate before market structure looks different.
AI can lower the cost of personalization, content production, experimentation, service, and software creation. Rivals may improve the economics of competing before incumbents recognize that the basis of competition has changed.
POLARIS QUESTION
What advantage still compounds if competitors gain similar intelligence?
Why this is a hypothesis, not a rewrite of Porter.
Porter's model describes the structural forces that shape industry profitability. Polaris Shift is testing whether AI is best understood as an independent force or as a cross-cutting source of leverage that changes the intensity and mechanics of the existing five. Until that case is demonstrated, this page uses the more conservative interpretation.
Research hypothesis
There is no single strategic half-life.
Strategy is built from assumptions that decay at different speeds. Purpose can remain useful for generations while interfaces, workflows, technology choices, and metric targets can lose relevance much faster. The question is not how often to change strategy. It is which assumption is expiring.
STRATEGIC HALF-LIFEHow quickly can environmental change invalidate the assumptions behind a strategic choice?
PURPOSELongest
CUSTOMER VALUE / JOBLong
POSITIONING / BUSINESS MODELMedium
DISTRIBUTION / TECHNOLOGYShorter
WORKFLOW / METRIC TARGETSShortest
PROXY DURABILITY BY CONTEXT
The relative clocks above describe strategic layers. The contexts below ask a narrower question: how quickly might the current measurable proxy lose meaning?
SHORTER HALF-LIFELONGER HALF-LIFE
ILLUSTRATIVE CONTEXT
Search and discovery
The human interface can move from query and click behavior toward synthesis, recommendation, and action. Interaction metrics can therefore lose meaning faster than the underlying purpose.
PROXY UNDER PRESSURE
Queries, clicks, sessions, time in interface.
REVALIDATION QUESTION
If useful outcomes happen with fewer searches and fewer clicks, what should replace interaction volume as evidence of value?
Strategic Half-Life and the proxy spectrum are Polaris Shift research hypotheses, not forecasts or universal time constants. Any real assessment should be grounded in a specific company, industry, customer value model, economics, and evidence.
Research exhibit
The Metric Graveyard
Most dangerous metrics are not useless. They became dangerous when organizations asked them to represent more value than they could safely carry.
THE RULEA proxy can be useful and still be the wrong North Star.
PROXYTime spent
LEGITIMATE USE
A useful behavioral signal when more time genuinely correlates with meaningful engagement.
FAILURE MODE
More time can mean delight, friction, confusion, addiction, or work that AI should have removed.
BETTER FORMULATION
Successful outcomes plus satisfaction, trust, and retention quality.
The graveyard is not a list of bad metrics. It is a reminder that every measure has a legitimate role, and every proxy can become misleading when the system around it changes.
Polaris Shift should be willing to discover that a signal is weaker, slower, narrower, or less structural than expected. Rigor matters more than proving the framework right.
Study public company purpose, vision, economic objectives, and AI-era pressure while keeping documented language separate from Polaris interpretation.
Research agenda · Planned
The State of the North Star
A future research program examining whether organizations are changing what they optimize for as AI changes their strategic environment.
Organizational study · Planned
Reorientation at Scale
How company stage, legacy depth, governance, system interdependence, sunk commitments, and customer expectations change the effort required to reorient strategy.
Signal study · Planned
Agentic Customer Journeys
Tracking where consumers are willing to delegate discovery, evaluation, decision-making, and action to AI, and what that changes for brands.
Operating model study · Planned
From AI Pilots to Organizational Rewiring
How organizations move from isolated experimentation toward distributed capability, bounded agency, and compounding institutional learning.
Field research · Planned
Polaris Shift Conversations
A repeatable interview and survey program across enterprise leaders, builders, and independent thinkers to test which assumptions are changing, what remains stable, and where organizations struggle to reorient.
Research principles
Prefer direct sources. Separate capability from adoption. Look for behavior, not announcements. Triangulate telemetry, human experience, and outcomes. Interrogate variance before trusting averages. Make counterevidence visible. Separate evidence from interpretation. Use confidence language when the evidence is incomplete.
Continue your course
Evidence becomes useful when it changes the question.
Move from research methodology into company interpretation, live signals, or your own strategic assumptions.