STRATEGYMonths to result

CIA Evidence-Based Pattern Assessment

Find accurate explanations by cataloging evidence patterns and resisting your assumptions.

Problem it solves

Professionals jump to emotionally satisfying conclusions about people or situations rather than finding the explanation that best fits all available evidence.

Best for

Leaders, HR professionals, and strategists who need accurate assessments of people, competitors, or situations where premature conclusions carry a material cost.

Not ideal for

Time-critical decisions where the cost of delayed action materially exceeds the cost of an imperfect assessment.

Overview

Why this framework exists

Intelligence tradecraft demands accuracy over speed and emotional satisfaction. This method trains practitioners to suspend initial gut reactions, catalogue specific observable behaviors across multiple contexts, generate all plausible explanations, and select the most parsimonious one consistent with all evidence while actively hunting for disconfirming data. Unlike intuition-based assessment, this method treats assumptions as adversaries. CIA operative Andrew Monte describes it as looking for the most explainable pattern that fits the evidence rather than confirming what you already believe—a discipline transferable to talent assessment, competitor analysis, and any situation where jumping to conclusions carries a cost.

Core principles

5 total
  1. Assumptions are the enemy of accurate assessment
  2. The most parsimonious explanation—not the most dramatic—is usually correct
  3. Evidence accumulates only if you stay in question long enough
  4. Behavioral data must be specific and observable, not interpretive labels
  5. Ruling out explanations is as valuable as identifying the right one

Steps

6 steps
  1. Externalize and suspend your gut reaction
    Write down your initial instinct about the person or situation, label it 'Hypothesis Zero,' and explicitly set it aside. It will be one of several explanations to evaluate—not the answer to confirm.
    Pro tipExternalizing the assumption removes its invisible authority over the analysis. You cannot challenge what you have not named.
  2. Catalog specific observable behaviors
    Document what the subject actually did, said, or decided across multiple situations using the format: '[Subject] did [specific action] in [specific context].' No interpretations yet—only observable facts.
    WarningAvoid collapsing behaviors into labels such as 'he was aggressive' before the generation phase. Labels are conclusions dressed as data.
  3. Generate all plausible explanations
    List every credible explanation for the behavioral pattern, including the most charitable and the most concerning. Do not evaluate or rank any of them yet—generation and evaluation are separate cognitive tasks.
    Pro tipForce a minimum of three distinct explanations before evaluating any. This breaks anchoring on your first plausible theory and opens the solution space.
  4. Apply the parsimony test
    Identify which explanation accounts for the most behavioral data points while requiring the fewest additional assumptions or invented variables to make it fit.
    Pro tipThe explanation that requires the least invention to accommodate the evidence is usually closest to the truth—this is behavioral Occam's Razor.
  5. Hunt for disconfirming evidence
    Deliberately look for any behavior, data point, or context that would disprove your leading explanation. If you cannot find any disconfirming evidence, you likely have not looked hard enough.
    WarningConfirmation bias is most dangerous at this stage. Treat finding disconfirming evidence as a success, not a setback—it is doing exactly what the method requires.
  6. Document your assessment with revision conditions
    Record your best current explanation along with the specific evidence it rests on and the threshold of new information that would cause you to revise it, before that new information arrives.
    Pro tipPre-committing to revision conditions protects against the sunk-cost trap of defending an assessment past its useful life simply because you publicly committed to it.

Checklist

Saved in your browser

Examples

2 cases
Behavioral Profiling of Kanye West

Rather than accepting the 'narcissist' label, analyst Andrew Monte catalogued specific behaviors across multiple clips: behavior that had changed over time, raw emotional responses rather than manipulation, instant de-escalation upon validation, and hyper-organized verbal planning. He generated narcissism, psychopathy, and developmental disorder as competing explanations, then applied parsimony—the developmental and mental health explanation required fewer additional assumptions to fit all the data.

OutcomeA more accurate behavioral profile that avoided a common misdiagnosis and identified more actionable strategies for those managing the relationship.
Lisa Bilyeu / CIA Spy: Is Kanye Mentally Ill or a Narcissist?
Competitor Price-Cut Assessment

A strategy team's gut reaction to a competitor's sudden price cut in one segment was 'they're trying to destroy us.' Applying the CIA method, they catalogued six behaviors—price cut limited to one segment, recent leadership change, flat marketing spend—and generated three explanations: cash flow stress, market testing, and defense against a new entrant. The most parsimonious explanation proved correct.

OutcomeThe team avoided a costly retaliatory price war and focused differentiation efforts against the actual threat rather than the imagined one.

Common mistakes

3 traps
Treating the first compelling explanation as final
Once a plausible explanation emerges, the brain stops generating alternatives. Force the complete list of competing explanations before evaluating any single one—generation and evaluation must be kept as separate sequential steps.
Using interpretive labels as behavioral data
'He seemed manipulative' is an interpretation, not a data point. Grounding every input in specific observable actions is what keeps the analysis valid, challengeable, and free from circular reasoning.
Skipping the disconfirmation step
Most people generate an explanation and then confirm it. The CIA method requires actively hunting for evidence that your leading explanation is wrong before committing—skipping this step produces confident but inaccurate assessments.

Origin story

How this framework came to be

Described by CIA operative Andrew Monte on the Lisa Bilyeu show as his core analytical training: 'I was taught not to trust my assumptions and instead to always pursue some kind of assessment.'

Source

Traced to primary
Source · VIDEO
CIA Spy: Is Kanye Mentally Ill or a Narcissist? This is What His Behavior REALLY Reveals... — Lisa Bilyeu
Lisa Bilyeu · 2026
Open source →

Related frameworks

Browse all Strategy →