AISFY — AI impact baseline worksheet Version: 21 October 2026 Guide: https://aisfy.ai/blog/ai-impact/ai-impact-measurement-baseline/ Purpose: Compare equivalent accepted work before and during a bounded AI trial. This is an editorial planning aid, not a validated measurement product. Use approved or fictional inputs. Do not include personal or confidential data. 1. AGREED COMPARISON Task / unit of work: Evidence owner and reviewer: Observation dates and sample size: Acceptance criteria (accuracy, completeness, sources, action boundaries): Comparable input complexity and participant experience: Approved tool and version / access constraints: 2. RECORD EACH ATTEMPT (copy this section for every baseline and trial attempt) Attempt ID and date: Method (baseline / AI-assisted): Preparation minutes: Drafting minutes: Review minutes: Rework minutes: Total minutes = preparation + drafting + review + rework: Accepted output? Yes / No (record failed attempts too): Quality errors and corrections: Tool / process costs, where known: Evidence reference (do not paste sensitive material): Unusual conditions / uncertainty: 3. REVIEW THE EVIDENCE Number of accepted and rejected outputs for each method: Typical effort and variation for comparable tasks: Quality differences and unresolved errors: What changed besides the AI tool or training? Decision: continue / adapt / stop: Decision owner and next review date: Limits of the conclusion: 4. FICTIONAL WORKED EXAMPLE — NOT AN AISFY CUSTOMER RESULT Baseline: preparation 20 + drafting 30 + review 10 + rework 0 = 60 minutes. AI-assisted: preparation 20 + drafting 10 + review 25 + rework 0 = 55 minutes. If both outputs pass the same quality standard, the difference is 5 minutes, or approximately 8.3% of baseline effort. Drafting alone improved by 20 minutes; that is not the total saving. This single example does not establish causation. Do not equate freed time with realized cash savings or ROI. Account for tool costs, setup, failed attempts and how time is actually redeployed separately. Training evaluation: https://aisfy.ai/blog/ai-impact/measure-ai-training-effectiveness/ Training options: https://aisfy.ai/ai-training/