Attribution models (last-click, multi-touch) in Digital Marketing - Mini Project: Build & Apply

Understanding Attribution Models: Last-Click and Multi-Touch

📖 Scenario: You work as a digital marketing analyst for an online store. Your manager wants to understand how different marketing channels contribute to sales. You will create a simple model to represent customer journeys and apply two common attribution models: last-click and multi-touch.

🎯 Goal: Build a step-by-step representation of customer journeys and apply last-click and multi-touch attribution models to assign credit to marketing channels.

📋 What You'll Learn

Create a dictionary called customer_journeys with exact customer IDs as keys and lists of marketing channels as values.

Create a variable called last_click_credit to store the last-click attribution results.

Create a variable called multi_touch_credit to store the multi-touch attribution results.

Use a for loop with variables customer and channels to iterate over customer_journeys.items().

Assign full credit to the last channel in each journey for last-click attribution.

Assign equal credit to all channels in each journey for multi-touch attribution.

💡 Why This Matters

🌍 Real World

Marketers use attribution models to understand which marketing channels help drive sales and how to allocate budgets effectively.

💼 Career

Digital marketing analysts and managers rely on attribution models to measure campaign performance and optimize marketing strategies.

1

Create customer journeys data

Create a dictionary called customer_journeys with these exact entries: 'C001': ['Email', 'Social Media', 'Paid Search'], 'C002': ['Direct', 'Email'], and 'C003': ['Social Media', 'Paid Search', 'Email'].

Digital Marketing

# Create the customer_journeys dictionary with exact data
# Your code here

Need a hint?

Use a dictionary with customer IDs as keys and lists of channels as values.

2

Set up attribution result containers

Create two empty dictionaries called last_click_credit and multi_touch_credit to store attribution results.

Digital Marketing

customer_journeys = {
    'C001': ['Email', 'Social Media', 'Paid Search'],
    'C002': ['Direct', 'Email'],
    'C003': ['Social Media', 'Paid Search', 'Email']
}
# Create empty dictionaries for attribution results
# Your code here

Need a hint?

Use empty dictionaries to store the attribution credits for each channel.

3

Calculate last-click and multi-touch attribution

Use a for loop with variables customer and channels to iterate over customer_journeys.items(). For last-click attribution, assign full credit (1) to the last channel in channels. For multi-touch attribution, assign equal credit (1 divided by the number of channels) to each channel in channels. Update last_click_credit and multi_touch_credit dictionaries accordingly, adding to existing values if the channel appears multiple times.

Digital Marketing

customer_journeys = {
    'C001': ['Email', 'Social Media', 'Paid Search'],
    'C002': ['Direct', 'Email'],
    'C003': ['Social Media', 'Paid Search', 'Email']
}
last_click_credit = {}
multi_touch_credit = {}
# Use a for loop to calculate last-click and multi-touch attribution
# Your code here

Need a hint?

Remember to add credit to existing values using dict.get(key, 0) to avoid overwriting.

4

Complete attribution model setup

Add a comment at the end of the code explaining that last_click_credit shows the total credit per channel using last-click attribution, and multi_touch_credit shows the total credit per channel using multi-touch attribution.

Digital Marketing

customer_journeys = {
    'C001': ['Email', 'Social Media', 'Paid Search'],
    'C002': ['Direct', 'Email'],
    'C003': ['Social Media', 'Paid Search', 'Email']
}
last_click_credit = {}
multi_touch_credit = {}

for customer, channels in customer_journeys.items():
    last_channel = channels[-1]
    last_click_credit[last_channel] = last_click_credit.get(last_channel, 0) + 1
    credit_per_channel = 1 / len(channels)
    for channel in channels:
        multi_touch_credit[channel] = multi_touch_credit.get(channel, 0) + credit_per_channel
# Add a comment explaining the attribution results
# Your code here

Need a hint?

Write a clear comment summarizing what each dictionary represents.