type
Post
status
Published
date
Nov 3, 2025
slug
casestudy
summary
tags
R
Excel
Tableau
category
Data
icon
password
📖
Here are some of the case studies I have finished

📝 1 Digital Analysis

1.1 Sentiment Analysis with Naïve Bayes classifier

 

1.2 A/B testing in online experiment

 

1.3 Recommendation System with KNN

 

1.4 SEO: help improve a local small business website

notion image
SEO is different from SMO. It is not a paid promotion but a natural promotion from the search engine.

📎 Reference

网站谷歌SEO入门指南网站谷歌SEO入门指南谷歌SEO学习指南 | 网站谷歌SEO入门指南
SE RankingSE RankingSE Ranking — Robust SEO Software for Every Major Task
 

1.4 Search Engine Marketing

Paid Search:
Metric
Key word
The keyword is important as it relates to where the ad is shown and how it is shown, directly influencing traffic, cost, and conversion.
—Impressions —Click-through-rate CTR
Big amount(max PCP)
Big amount(max PCP) is important as it determines the traffic and costs. Take an example of the following to understand how to calculate the cost:
Clicks and Cost-per-click (CPC)
Ad Copy
Ad Copy is important
Transaction Coversion Rate (TCR)
Land Page
Land Page is important
Conversion
notion image
  1. Ad Rank = Max CPC × Quality Score, which determines the ad order.
  1. Actual CPC = (Ad Rank of advertiser below) / (Your Quality Score) + 0.01(optional).
  1. Here, D is in position 2, next is B (Ad Rank=0.9), D's Quality Score=1, so CPC ≈ 0.9.
 
PPC with Google Ad
 

📝 2 Pricing Analytics and Strategies

In this class, we complete the case study and then listen to how the standard process should be, so that we can learn from our mistakes.

2.1 Atlantic Computer

Choose from four pricing options:
1)Status quo: stick with the tradition by not charging for PESA software
2)Competition-based: sell for the equivalent of the competitor's best alternative.
3)Cost-plus: fixed profit margin
notion image
4)Value-based pricing
Class Note on the calculation of TEV
notion image
 
 

2.2 Xbox

The price discrimination: giving a different price for different segmentation
Key takeaway: the subscription model of Xbox, what kind of pricing to be used: usage-based loyalty loyalty-based, or both.
 

2.3 Metabical

notion image

2.4 Behavioral Pricing

 

2.5 Pink Tax

How the pink tax could work: women and children are more willing to spend more based on the packaging instead of the functional value.
Takeaway: Be price-sensitive and be aware of pink tax.
 

2.6 Eat App-Bundling

Class Notes:
notion image
 
notion image
notion image
 
notion image
Takeaway: A menu of subscription plans to serve heterogeneous consumers. A free tier can be provided with limited features. Upgrading and cross-annual discount increases retention.
 

2.7 Bundling Case - 3 scenario

notion image

2.8 Porsche: the subscription model of car

 

2.9 Netflix: the subscription-pricing model for online service

 

2.10 Computron: B2B pricing

 

2.11 Jellycat pricing: optional essay

 

📝 Customer Relationship

As part of my program, I’ve been diving into a series of case studies that examine how companies approach customer relationships and customer relationship management (CRM). One of the courses where this really came to life is Customer Relationships Analysis and CRM with Professor Daniel.

Case Study 1: Amazon and Customer Centricity

We began with Amazon as an example of a customer-centric strategy. The company has built its reputation on using data-driven tools—particularly its Customer Relationship Value (CRV) algorithm—to refine its customer interactions continuously. What’s fascinating is how Amazon continually shifts and reframes its approach to keep customer satisfaction at the center, adapting its CRV models over time.
To contrast this, we examined Costco, which has achieved remarkable customer loyalty without heavily segmenting or overpersonalizing. Instead, Costco keeps things simple with a membership model and consistent value delivery. Both Amazon and Costco are deeply customer-centric, but their strategies reveal very different ways to execute that philosophy.

Case Study 2: MoviePass

Next, we turned to the story of MoviePass, a company that launched in 2011 with the ambitious idea of offering unlimited movie tickets for a flat monthly fee. Despite attracting attention, MoviePass struggled with sustainability and eventually shut down in 2019 amid financial and legal challenges.
What makes this case so interesting is its relaunch in May 2023. With adjustments to its business model—including a more sustainable contribution margin estimated at $40 per day—the company is now positioning itself for a much stronger future. This new approach suggests that MoviePass has learned from its past missteps, and it could very well be a success this time around.
 

Exploring Customer Lifetime Value (CLV)

Another important concept we explored was Customer Lifetime Value (CLV). There are several different ways to calculate CLV, and each method comes with its own limitations.
One example we studied was the finite horizon model, which estimates CLV within a set timeframe. This framework highlights how customer value can evolve as businesses grow—something we’ve seen in practice with platforms like ChatGPT, whose user base and monetization strategies continue to shift over time.
We also looked at undiscounted CCLA (Customer Contribution Lifetime Analysis) through the case of a company called Plainton, using insights from the article How to Value a Company by Analyzing Its Customers by Daniel McKenzie and Peter Fader. This exercise showed how businesses can be valued not just on revenue streams but on the long-term worth of their customers.

Customer Acquisition Costs (CAC)

From there, we transitioned into the Customer Acquisition course, where we examined how companies manage the cost of bringing in new customers.
A case study of Form SF1 illustrated the dynamics of extreme spending on customer acquisition. While its acquisition costs decreased over time, we also observed how they eventually began to rise again, showing the cyclical nature of CAC management.
To deepen our understanding, we analyzed Facebook Ads and the challenges of accurately calculating acquisition costs. We noted that certain elements should not be included in CAC calculations, such as:
  • Sales marketing costs for repetitive orders
  • True cost of goods
  • Fixed overhead
This distinction is critical for ensuring the numbers reflect the actual efficiency of customer acquisition strategies.
For platforms like Facebook or LinkedIn, the rule of thumb is to maintain an LTV-to-CAC ratio of at least 3:1. This ensures that the long-term value customers bring in is more than enough to cover the costs of acquiring them, keeping the company’s growth financially healthy.
While CAC tends to increase over time, well-managed companies work hard to prevent the LTV/CAC ratio from deteriorating as they scale.
 

Case Study 3: Auditing a Customer Base

In the course Audit as a Customer Base, we studied how to evaluate the health and sustainability of a company’s customer-driven business model.
Our case study focused on Rent the Runway, an online rental service for designer clothing and accessories. Its model is built around subscription-based fashion reuse, enabling customers to rent rather than buy items.
Key milestones for Rent the Runway include:
  • 2009 – Founded
  • 2016 – Launched its subscription service
  • Author:Yunzhu HUANG
  • URL:/article/casestudy
  • Copyright:All articles in this blog, except for special statements, adopt BY-NC-SA agreement. Please indicate the source!
From Python Maps to Web Deployment: Building Interactive Data VisualizationsMy educational journey
Loading...