__STYLES__
Tools used in this project
Cafe Reward Promotion Program

About this project

I worked this project as per below steps.

  1. Define Project Objective

Maven Cafe conducted the test offering different combinations of promotion to existing rewards members. The test has conducted for 30-day period. Based on that test, this project is to identify key customer segments and develop a data-driven strategy for future promotional messaging & targeting.

  1. Data Exploration

I used Python to understand about the variables in the tables.

  1. Data Wrangling

In this step, I worked for missing values, outliers and data enrich. 13% of total records are missing in gender and income field which is a lot to remove from the dataset. Therefore I used backfill method to impute the value. After I have checked, this imputation does not cause bias. undefinedI used IQR method to find and fix the outliers.

After that all 3 tables are merged and adding additional columns to support the analysis. After that I export the merged file to be used in Tableau.

  1. Visualization Building

  2. Summary

Below is the finding and recommendation after the dashboard.

Key Insight

  • Maven Cafe conducted promotion program to 17,000 members. Revenue for 30 days period from those members is $1.78 M. Reward back to member is $0.16M which is 9% of revenue.

  • Members equally contributed to revenue in terms of gender wise. Middle Age and Senior Adult are the most revenue contributor with over 90%. Middle income group members take 73% revenue contribution. Member with less than 10 years membership tenure is with 99% of contribution.

  • There are 76,277 promotion offered to members with (info:20%, bogo: 40%, discount : 40%). 75.7% of promotion were viewed and 55% of (bogo and discount) promotion were completed. More than 90% of promotion offered via all 4 available channels are viewed by member.

  • Although there are awareness by members about promotion, number of completed promotion is still low with 55%.

Recommendation

  • Key members for Maven Cafe are Middle Age and Senior Adults with Middle income range and less than 10 years membership people whose revenue contribution is more than 90%.

  • To get more awareness from members on promotion, all 4 available channels should be used. Which make sure to get more than 90% of member viewed.

  • Social channel is the most important one which can help to get at least 80% of member viewed.

  • As per the analysis, low promotion completed rate happened due to below points

  1. Promotions were offered to non-active members who do not have any transactions within 30 days

  2. Some active members can not complete the promotion due to not enough of minimum amount

  3. Some active members can not complete the promotion due to exceed of promotion duration

  • Coming promotion program should focus to offer the promotion based on the member past transactions and should be offered to more frequently visited members to Cafe.

Additional project images

Discussion and feedback(4 comments)
comment-1661-avatar
Murilo Evangelista
Murilo Evangelista
23 days ago
Hi Cherry, Well done. Have some questions 😊 How could you "calculate" membership tenure without more specific information about períod? And also about the channels, there is no specification about the channel on the events, I mean, how can I know if the offer arrives from social or e-mail for a specific customer? Maybe I´m missing something (🤔😅). Many thanks, Murilo

comment-1665-avatar
Walid Hussein
Walid Hussein
21 days ago
Nice job, Cherry! Hi Murilo, for the membership tenure, you can use the "Became Member On" column in the customer table to get this information. Also, you can split the marketing channels in the offer table to get binary values for each channel. That's what I did.

comment-1666-avatar
Cherry Aye Mya Mya Tun
Cherry Aye Mya Mya Tun
Project owner
21 days ago
Project owner
Hi @Walid Hussein, thanks for answering to Murilo. Yes, this is the same that I did for membership tenure and channels.

comment-1681-avatar
Murilo  Evangelista
Murilo Evangelista
17 days ago
Hi Walid, thanks for the answer. But you know, I mean there is no clear information to take these assumptions (my point of view). I understand what you´ve done, but for example, the offer id: ae264e3637204a6fb9bb56bc8210ddfd (bogo) there are 3 different channels, did the customer receive it on all of them? Which one he/she used to see and converted in a transaction? 🤔 And about the tenure I got it (if we consider that these transactions occurs in 2024 😁).
2000 characters remaining
Cookie SettingsWe use cookies to enhance your experience, analyze site traffic and deliver personalized content. Read our Privacy Policy.