# Sourcing Analytics and Channel Performance: practice workbook
Talent Engineering Academy | An education initiative by Vitae
Edition 2026.09 | Updated 2026-09-24
Course: https://talentengineering.org/courses/sourcing-analytics

Measure channel quality, understand attribution and make sourcing decisions from comparable evidence.

## Your deliverable
A channel scorecard with cohort definitions and an experiment plan.

Use fictional information. Keep your completed work in your own secure notes. Exercises and capstone work are self-directed, not independently assessed.

## 1. Funnel event definitions

Your notes:



## 2. Cohort and observation window

Your notes:



## 3. Channel attribution rule

Your notes:



## 4. Conversion, time and cost calculations

Your notes:



## 5. Quality and experience guardrails

Your notes:



## 6. Interpretation and next experiment

Your notes:



## Lesson exercises

### 1. Define a sourcing funnel you can trust

Define six funnel events and a cohort rule for a fictional four-week campaign.

Your response:


Worked example: A report divides this week’s hires by this week’s messages even though the hires came from earlier searches. The revised report follows outreach cohorts and states the maturity window for each.

Self-review guidance: Include event timing, duplicate handling and incomplete outcomes. Explain why a recent cohort may still be open. A coherent funnel follows related events rather than mixing unrelated weekly totals.

### 2. Calculate conversion, effort and cost

Calculate interest rates and cost per interested conversation for two fictional channels, including recruiter time.

Your response:


Worked example: Channel A produces 8 interested replies from 40 contacts, or 20%. B produces 3 from 10, or 30%. B’s higher observed rate is based on a much smaller sample and does not establish a durable advantage.

Self-review guidance: Show counts, formulae and assumptions. If a channel has no interested conversations, report the cost and zero outcome rather than inventing a finite cost-per-outcome value. Explain sample uncertainty.

### 3. Handle attribution and selection effects

Assign credit under two attribution rules for three fictional multi-touch hires.

Your response:


Worked example: A candidate is first found at an event, later nurtured by email and finally applies through the website. A last-touch report credits the website; the relationship history shows why that is incomplete for budget decisions.

Self-review guidance: Explain how the totals change and which decision each rule supports. Do not call either rule proof of causation. Preserve the underlying journey so the team can interpret the allocation.

### 4. Choose the next sourcing experiment

Design a two-week experiment and specify what result would change your current plan.

Your response:


Worked example: A channel has many replies but few relevant conversations. The next test clarifies role scope in the first message, tracking informed interest and complaints rather than simply increasing send volume.

Self-review guidance: Define comparable groups or state why comparison is limited. Include a stop condition, observation window and decision owner. Predefine the interpretation so the team does not choose a success metric after seeing results.

## Portfolio review
Check that your work is internally consistent, distinguishes facts from assumptions, names decision owners and explains its limitations. Revise gaps before using the method in real work.

## Further reading
- [CIPD: Recruitment](https://www.cipd.org/uk/knowledge/factsheets/recruitment-factsheet/): Professional context for the recruitment lifecycle.
- [ICO: Recruitment and selection](https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/employment/recruitment-and-selection/): UK guidance. Check its current status and updates before implementation.

Original educational scenarios. References provide further reading and do not imply endorsement. Check current official rules and appropriate professional advice for real legal, financial or regulated decisions.