AI Content Performance Analyst Workbook
A printable workbook for completing the Content Performance Analysis and Experiment Pack across 7 course sections.
Download the AI Content Performance Analyst Workbook
Use this workbook with Build an AI Content Performance Analyst. Complete one numbered section per module and assemble the final Content Performance Analysis and Experiment Pack.
What the workbook contains
- Choose the Decision and Performance Objective
- Build the Content Data Contract
- Define Comparable Performance Measures
- Analyze the Account's Own Content
- Add Competitor or Adjacent Context Carefully
- Turn Findings into Ten Content Experiments
- Test and Operate the Analysis Loop
Each section gives you fields for the decision, evidence, metric rule, owner, test result, limitation, and next action required by its matching lesson. The final pages combine those records into the capstone and acceptance review.
How to use it
Start with one account, one platform, one observation window, and one business decision. Complete the matching workbook section after each lesson rather than filling the document from memory at the end. Keep raw rows, denominators, paid status, observation dates, and missing fields visible. Separate observed patterns from hypotheses. When a metric or comparison test fails, preserve the failed result, record the repair, and rerun the affected case.
Review the completed pack with the person who owns the content decision. A polished analysis is not enough: eligibility rules, missing-data treatment, comparison limits, experiment measures, and stop conditions must agree with the course acceptance criteria.
Important boundaries
The workbook is a planning and testing aid. It does not prove causation, future performance, ranking improvement, production readiness, provider availability, or revenue impact. It does not authorize publishing, paid distribution, data collection, account access, outreach, or deployment. Those actions require current evidence and separate approval.
Use the synthetic fixture only to learn the method. Replace it with permitted real or sanitized data before making a business decision.
Pick one content decision to investigate
The nine-page PDF helps structure a review; it does not connect to your analytics accounts. Choose a question such as which articles deserve an update, then write the reporting period, content set and action the analysis could support. Keep traffic, engagement and commercial outcomes as separate measures.
As a fictional exercise, compare an older article with a newer campaign page. Record publication dates, distribution and paid promotion before interpreting their totals. More visits may reflect more exposure rather than stronger content. A page with little traffic may not yet have enough evidence for a confident conversion comparison.
Write a recommendation someone can check
For each proposed action, include the observation, its source and the uncertainty that could change the decision. “Refresh the introduction because readers cannot find the promised answer” needs page evidence; “this topic is bad because it received fewer visits” skips several possible explanations.
Have a person approve changes to live copy, campaigns or budgets. Keep the original reporting extract and the decision so the next review can assess what changed. Continue with the content performance analyst course to practise moving from a report to a defensible editorial decision.