Project 01BeginnerBusiness4-6 focused hoursRetention analysis case study
LearnPracticeBuild

Customer Churn Analysis

Analyze customer behavior and identify the patterns most strongly associated with churn risk.

PythonPandasEDAData Visualization
FRAMEDATAANALYZEMODELEVALUATECOMMUNICATE

BUSINESS CONTEXT

Why this project matters

A subscription business is losing customers faster than expected. The retention team needs a clear view of which customer segments are churning, what behaviors show early warning signs, and where intervention efforts should focus first.

PROJECT OBJECTIVE

What you are expected to accomplish

Investigate the customer base, calculate churn patterns across meaningful segments, and turn the analysis into practical retention recommendations that a business stakeholder could act on.

DATASET OVERVIEW

Expected dataset structure

Expect a customer-level table covering subscriptions, billing behavior, support activity, and whether the customer churned within a recent evaluation window.

Main entities / rows

One row per customer account

Target variable

churned (Yes/No)

customer_id

string

Unique identifier for each customer account.

tenure_months

integer

Number of months the customer has been active.

contract_type

categorical

Monthly, yearly, or multi-year contract plan.

monthly_charges

float

Current recurring subscription charge.

support_tickets_last_90d

integer

Recent support interactions that may signal friction.

auto_pay

boolean

Whether the customer uses automatic payment.

internet_service

categorical

Primary service package used by the customer.

churned

binary

Target flag indicating whether the customer left.

Data quality issues to expect

  • Tenure or billing fields may contain missing values for newer accounts or migration cases.
  • Categorical service labels may be inconsistently capitalized across source systems.
  • A few customers may have duplicate records because of plan changes or billing merges.

QUESTIONS TO ANSWER

Focus the work around meaningful decisions

  1. 01

    Which contract types show the highest churn rate?

  2. 02

    How does churn differ across tenure bands and monthly charge levels?

  3. 03

    Are customers with more recent support interactions churning at a higher rate?

  4. 04

    Which service combinations or customer segments appear most retention-sensitive?

  5. 05

    What practical actions could reduce churn in the highest-risk segments?

PROJECT ROADMAP

Move through the case study in a practical sequence

  1. 01

    Step 01

    Understand the retention goal

    Clarify how churn is defined and which business decision the analysis will support.

  2. 02

    Step 02

    Inspect the customer table

    Check record counts, data types, missingness, and whether each row truly represents one customer.

  3. 03

    Step 03

    Clean the dataset

    Resolve duplicates, standardize categories, and handle missing values in a documented way.

  4. 04

    Step 04

    Profile churn with EDA

    Measure overall churn and compare it across contract, tenure, billing, and service segments.

  5. 05

    Step 05

    Build useful segment features

    Create tenure bands, charge buckets, or service-count features that sharpen the analysis.

  6. 06

    Step 06

    Optional baseline modeling

    If you choose, test a simple classifier to compare feature importance against your EDA findings.

  7. 07

    Step 07

    Communicate business recommendations

    Summarize the strongest churn patterns and suggest realistic retention actions.

TASKS / MILESTONES

Concrete work to complete

  • Calculate the overall churn rate and confirm the analysis period.
  • Inspect missing values and document which fields need cleaning before analysis.
  • Group churn by contract type, tenure band, and payment method.
  • Compare monthly charges for churned versus retained customers.
  • Analyze churn by support-ticket frequency and service package combinations.
  • Create 2-3 segment features that make the patterns easier to explain.
  • Write final retention recommendations tied to specific segments or behaviors.

SUGGESTED VISUALIZATIONS

Visuals worth creating

  • Churn rate by contract type bar chart
  • Tenure distribution split by churn status
  • Monthly charges box plot by churn status
  • Service-segment comparison chart
  • Heatmap of churn rate across tenure and charge bands

DELIVERABLES

What the learner should produce

  • Cleaned analysis-ready customer dataset
  • Notebook or report documenting the workflow
  • 5-8 useful visualizations tied to churn questions
  • Segment-level churn insights
  • Final retention recommendations in plain business language

SUCCESS CRITERIA

What strong completion looks like

  • The project clearly defines churn and the grain of the dataset.
  • The analysis identifies meaningful churn differences across customer segments.
  • Cleaning choices and assumptions are documented rather than hidden.
  • Recommendations follow directly from the evidence in the analysis.

GUIDANCE

Support when you need a nudge

Reveal practical hints, checkpoints, and framing help without exposing a full finished solution.

Complete this project when the work is yours

Use this only after you have worked through the roadmap, milestones, and your own analysis.

KEY LEARNING OUTCOMES

Skills reinforced by this project

  • Translate a retention question into a structured analysis workflow.
  • Use EDA to find actionable customer-risk patterns.
  • Handle missing values and noisy categories in a business dataset.
  • Present churn insights in terms a non-technical team can use.

RELATED LESSONS

Revisit published lessons that support this build