Maven Cafe Card Rewards Program Report

Driving Customer Engagement Through Data

Challenge Overview

In this Maven Cafe Rewards Challenge, I assumed the role of a Sr. Marketing Analyst tasked with evaluating a 30-day promotional test. During this period, various combinations of offers were sent to the cafe’s rewards members, and the goal was to assess customer engagement, identify key customer segments, and propose a data-driven strategy to optimize future promotions.

As the test concludes, it’s essential to present the results to Maven Cafe’s Chief Marketing Officer (CMO), summarizing both customer behaviors and the effectiveness of the promotional strategies used.

The Objective

My mission as Sr. Marketing Analyst was two-fold:

1. Identify Key Customer Segments: Determine which customer groups were most (and least) engaged with the different promotional offers.
2. Develop a Data-Driven Strategy: Use insights gathered from the test to recommend a strategic approach for future promotional messaging and offer targeting, ensuring increased customer engagement.

The Dashboard Design Inspiration

The design was inspired by data visualization best practices, including the Bay Admin Multipurpose Dashboard By Code Grape (available on Behance here).

About the Data Set

The analysis is based on data simulating customer behavior over a 30-day period, capturing transactions and interactions with different promotional offers. The dataset is split into three key tables:

• Offers table: Information on the different promotional offers, including type, difficulty, and rewards.
• Customers table: Age, gender, income, and membership details of the cafe’s rewards members.
• Events table: Capturing how customers engaged with offers — whether they received, viewed, or completed an offer, and whether they made a transaction related to it.

For a transaction to be tied to an offer, it must occur at the same time the offer was “completed” by the customer. This nuanced behavior offers an important insight into how customers interacted with the promotions.

Datasets BEFORE cleaning

Data Cleaning Process

Before diving into the analysis, it was crucial to prepare the data. Some of the key cleaning steps included:

• Handling Missing Demographic Data: Upon inspection, the customers.csv file had missing values in fields such as age, gender, and income for over 2,000 customers. These records were flagged as incomplete and excluded from the main analysis to prevent skewing results related to segmentation and offer performance.
• Offer Timing & Duplication: Customers sometimes received and completed the same offer multiple times. Ensuring that each completion was attributed to the correct offer required cleaning and filtering the events data.
• Splitting the Value Column: The events.csv file had a Value column containing mixed data, including offer IDs, transaction amounts, and rewards. To ensure clearer analysis, this column was split into three separate columns:
• Offer ID: The identifier for the offer associated with the event.
• Amount: The transaction amount recorded during the event.
• Reward: The reward earned for completing an offer. This allowed for a cleaner and more precise analysis of customer transactions and offer-related rewards.
• Transaction Attribution: Transactions were only considered “completed” when they coincided with the completion of an offer. Properly linking these transactions was crucial for an accurate analysis of offer performance.
• Handling the “Became a Member On” Column: In the customers.csv dataset, the Became a Member On column recorded the date a customer joined the rewards program. The column was converted from text to a proper date format.
• racking Offer Timing: To measure customer responsiveness, we built calculated columns in Power BI, such as “Time to Complete Offer,” which tracked the duration between offer receipt and completion. Each “Offer Completed” event was correctly linked to the most recent “Offer Received” event for accurate offer-attribution analysis.

Key Insights

1. Customer Demographic Insights

Understanding the demographic breakdown of Maven Cafe’s customers offers a glimpse into how different groups interacted with the promotions:

Customer segment analysis page

• Female customers had a higher completion rate (58%) compared to male customers (44%)
• Higher-income customers ($75k+) were significantly more likely to complete offers
• Customers aged 56–70 were the most engaged, completing 52.70% of the offers they received
• Newer members (those who joined between 2017 and 2018) had a notably lower completion rate (42.57%) compared to more established members (65.49%)

2. Offer Insights

Offer insights page

Analyzing offer engagement provides insights into what types of promotions work best for Maven Cafe’s customer base:

• The overall offer completion rate was 49.88%
• Discount offers had completion rates ranging from 52.73% to 76.56%, whereas BOGO offers had lower rates (51.04% to 63.98%)
• Customers received between 1 and 7 offers throughout the 30-day period, and a few customers did not receive any offers at all.
• Offers with higher difficulty levels and shorter completion windows underperformed compared to others

Additional Images

Recommendations for Future Promotional Messaging & Targeting

Based on the insights gathered, the following data-driven strategies are proposed to optimize future promotional campaigns:

1. Target High-Value Segments:

Focus promotional efforts on high-value customer segments like high-income customers ($75k+), core members (joined between 2015 & 2016), and older age groups. These segments were the most engaged, and targeting them with tailored, premium offers could drive even greater participation.

2. Simplify Offers for Low-Engagement Groups:

Create more accessible offers for low-income (<$54k) and younger customers (under 25), who had lower completion rates. Simpler offers with smaller rewards may appeal more to these groups and increase engagement.

3. Improve Offer Distribution Strategy:

Shift from random offer distribution to a more strategic approach. Ensure customers receive a balanced number of offers throughout the campaign, and avoid sending too many offers to prevent fatigue. This balanced approach can lead to more consistent engagement across the entire customer base.

4. Leverage Multi-Channel Delivery:

Maximize reach by delivering offers through a mix of Web, Email, Mobile, and Social channels. Utilizing all distribution channels would greatly increase Offer Completion rates.

5. Prioritize Discount Offers:

Since Discount offers consistently outperformed BOGO (Buy One Get One) offers, Maven Cafe should prioritize these types of offers in future promotions. BOGO offers may still be valuable but should be simplified or paired with additional incentives to boost appeal.

6. Optimize Offer Frequency:

Spread offers more evenly throughout the promotional period, ideally every 5–7 days, to avoid overwhelming customers and ensure they remain engaged. Tailor frequency based on individual customer behaviors, allowing for more personalized and effective outreach.

Link to Dashboard:

Explore the full interactive dashboard here to dive deeper into the data and visualize customer engagement, offer performance, and the recommendations for promotional messaging & targeting.

Conclusion

The Maven Cafe Rewards Program holds great potential for boosting customer engagement and loyalty. By pinpointing key customer segments and analyzing how different offers appeal to each group, Maven Cafe can fine-tune its promotional strategy. The recommendations provided offer a clear path to better targeting, improved completion rates, and higher customer satisfaction.

In this experiment, my role as a Senior Marketing Analyst has set the foundation for a more strategic, data-driven approach to promotional messaging. These insights will help Maven Cafe turn analysis into actionable steps for future success.

Tools Used

• Data Cleaning: Microsoft Excel, Power BI.
• Dashboard: Power BI.
• Background Design: Figma.