COMPEL Glossary / baseline
Baseline
A baseline is a documented measurement of current performance, capability, or conditions taken before an AI initiative begins, providing the reference point against which progress, improvement, and ROI are measured.
What this means in practice
Without baselines, transformation success is subjective -- a matter of narrative rather than evidence. The COMPEL Value Realization Layer defines four baseline categories: process baseline (throughput, cycle time, error rates), cost baseline (comprehensive current cost structure), quality baseline (defect rates, satisfaction scores), and time baseline (end-to-end cycle times, time-to-decision). Baselines must be established using process mining, time-motion studies, or operational data analysis during the Calibrate stage -- establishing them retrospectively after deployment is significantly less reliable and may not be accepted by auditors.
Why it matters
Without documented baselines, transformation success is subjective — a matter of narrative rather than evidence. Baselines provide the reference point against which progress, improvement, and ROI are measured, transforming vague claims of improvement into quantifiable, auditable results. Establishing baselines retrospectively after deployment is unreliable and may not be accepted by auditors, making pre-deployment measurement essential.
How COMPEL uses it
Baselines are established during the Calibrate stage using process mining, time-motion studies, or operational data analysis. COMPEL defines four baseline categories: process, cost, quality, and time. These baselines anchor the measurement framework designed during Model and are the reference against which the Evaluate stage measures actual improvement. The Learn stage reviews baseline methodology to improve measurement accuracy in subsequent COMPEL cycles.
Related articles in the Body of Knowledge
Related Terms
Other glossary terms mentioned in this entry's definition and context.