COMPEL Glossary / value-thesis
Value Thesis
A value thesis is a testable hypothesis articulating the causal logic connecting an AI initiative to expected business outcomes.
What this means in practice
It follows the format: 'If we deploy [capability] in [workflow], then [metric] will improve by [amount] within [timeframe], because [mechanism].' Value theses must be specific, falsifiable, and subject to revision as evidence accumulates. They force practitioners to make their assumptions explicit rather than hiding behind vague claims of 'AI-driven improvement.' In the COMPEL framework, the Value Thesis Register (TMPL-C-006) is a mandatory Calibrate-stage artifact that documents hypotheses for each prioritized use case. These hypotheses are tested against actual outcomes during the Evaluate stage, creating a closed loop between planning and measurement.
Why it matters
Value theses force practitioners to make assumptions explicit rather than hiding behind vague claims of AI-driven improvement. The testable hypothesis format connects specific capabilities to measurable business outcomes with defined timeframes and causal mechanisms. Without value theses, organizations invest in AI initiatives that cannot be evaluated for success or failure, making it impossible to learn what works and allocate resources effectively.
How COMPEL uses it
The Value Thesis Register (TMPL-C-006) is a mandatory Calibrate-stage artifact documenting hypotheses for each prioritized use case. These hypotheses follow the format: 'If we deploy [capability] in [workflow], then [metric] will improve by [amount] within [timeframe], because [mechanism].' The Evaluate stage tests theses against actual outcomes, creating a closed loop between planning and measurement that drives evidence-based decision-making.
Related Terms
Other glossary terms mentioned in this entry's definition and context.