AB testing in Marketing Banners
Nara Nguyen
A. Project Objective:
Our company offer the discount price on subscription for whitelist group of users. They would like to know if we change the way the show the discount price, whether the conversion rate of users will be higher? (the original price of the subscription is 15$)
Content A: show content "You will buy this subscription only 10$"
Content B: show content "You're discount 5$ for this subscription"
B. Methodology:
1. Hypothesis Statements:
When users see the content A, they will tend to buy the subscription than content B.
2. Evaluate Metrics:
Primary metrics: Conversion Rate = Users successfully buy the subscription over the total users were offered
Secondary metrics: The number of users fail to buy over the total users were offered (users can have error in payment process - this metrics can support the main hypothesis)
3. Sample size & Duration:
Sample size: 27000 users with:
Alternative H1 is for: Equivalent
Confidence level: 95%
Power: 80%
Duration: the AB testing was conduct within 1 month
4. External Factors to monitor:
Payment Error: daily trend of conversion rate. If the conversion rate suddenly low, it can be caused by payment error by our system
Seasonal Sales
B. Results:
Control Group: Content B
Treatment Group: Content A
Conversion rate of users in Treatment group is: 36.43 %
Conversion rate of users in Control Group is: 34.69 %
P-value = 0.0031359111241115773
The result was statistic significant. We can use the content A for users.
Like this project
Posted Aug 29, 2024
Nara built an interactive dashboard to track and visualize key sales metrics for an e-commerce business, improving their decision-making process.
Likes
0
Views
0
Tags
