From Data to Business Impact

One of the biggest transitions from Data Analyst to Data Scientist is learning that companies do not care about models—they care about business outcomes.

A perfect machine learning model that increases revenue by $0 is less valuable than a simple dashboard that helps generate $1 million in sales.

This lesson introduces the language of business, marketing, and growth that every modern Data Scientist must understand.

Learning Objectives

By the end of this lesson, you should be able to:

• Understand the marketing funnel
• Understand customer acquisition and retention
• Define important marketing metrics
• Calculate Customer Acquisition Cost (CAC)
• Calculate Conversion Rate
• Calculate Return on Advertising Spend (ROAS)
• Calculate Customer Lifetime Value (LTV)
• Understand churn
• Connect statistics to business decisions

Why Marketing Analytics Matters

Consider a company that spends:

• $100,000 on advertising
• Acquires 500 customers

Management wants answers:

• Was the campaign successful?
• Which customers should we target?
• Which channels work best?
• How much can we afford to spend on advertising?

These are data science questions.

The Customer Funnel

Most businesses think of customers as moving through stages:

Awareness → Interest → Consideration → Purchase → Retention → Advocacy

Example:

• 100,000 people see an advertisement
• 10,000 click
• 2,000 visit the website
• 500 purchase
• 300 remain active customers
• 100 become loyal advocates

The objective of marketing analytics is to understand where customers drop out and how to improve movement through the funnel.

Funnel Conversion Rates

Suppose:

• 100,000 impressions
• 10,000 clicks

The click-through rate is:

$$CTR=\frac{10,000}{100,000}=0.10=10\%$$

Only 10% of viewers clicked the advertisement.

Business Question

Imagine two advertising channels.

Channel A

• 100,000 impressions
• 5,000 clicks

Channel B

• 100,000 impressions
• 12,000 clicks

The click-through rate formula is:

$$CTR=\frac{Clicks}{Impressions}$$

For Channel A:

$$CTR_A=\frac{5000}{100000}=5\%$$

For Channel B:

$$CTR_B=\frac{12000}{100000}=12\%$$

At first glance, Channel B appears better.

However, clicks do not generate revenue.

Purchases generate revenue.

This distinction is critical for data scientists.

Marketing Funnel Metrics

Every data scientist should know the following metrics:

Click-Through Rate (CTR)

CTR measures how often users click an advertisement.

Formula:

$$CTR=\frac{Clicks}{Impressions}$$

Example:

$$CTR=\frac{8000}{200000}=0.04=4\%$$

Interpretation:

Four percent of viewers clicked the advertisement.

Conversion Rate

Conversion rate measures how many visitors complete a desired action, usually a purchase.

Formula:

$$Conversion\ Rate=\frac{Purchases}{Visitors}$$

Example:

• 5,000 visitors
• 250 purchases

$$Conversion\ Rate=\frac{250}{5000}=0.05=5\%$$

Interpretation:

Five percent of visitors became customers.

Customer Acquisition Cost (CAC)

Customer Acquisition Cost measures how much money is spent to acquire one customer.

Formula:

$$CAC=\frac{Marketing\ Spend}{New\ Customers}$$

Example:

• Marketing spend = $50,000
• New customers = 200

$$CAC=\frac{50000}{200}=250$$

The company spends $250 to acquire each customer.

Why CAC Matters

Suppose:

$$CAC=250$$

If the average customer spends only $100, the company loses money.

However, if the average customer spends $2,000, the company is profitable.

Therefore, CAC must always be compared with customer value.

Customer Lifetime Value (LTV)

Customer Lifetime Value estimates how much revenue a customer generates during their relationship with a business.

A simple formula is:

$$LTV=Average\ Revenue\ Per\ Period\times Average\ Customer\ Lifetime$$

Example:

• Monthly spending = $50
• Average lifetime = 24 months

$$LTV=50\times24=1200$$

Expected customer value is $1,200.

LTV vs CAC

One of the most important relationships in business is:

$$LTV>CAC$$

Example:

$$LTV=1200$$

$$CAC=250$$

This is generally a healthy business model.

Now consider:

$$LTV=120$$

$$CAC=250$$

The company loses money on every customer acquired.

Return on Advertising Spend (ROAS)

ROAS measures the effectiveness of advertising.

Formula:

$$ROAS=\frac{Revenue}{Advertising\ Cost}$$

Example:

• Revenue = $500,000
• Advertising Cost = $100,000

$$ROAS=\frac{500000}{100000}=5$$

Interpretation:

Every $1 spent on advertising generated $5 in revenue.

Churn

Churn measures customer loss.

Formula:

$$Churn=\frac{Customers\ Lost}{Customers\ at\ Start}$$

Example:

• Starting customers = 1,000
• Lost customers = 80

$$Churn=\frac{80}{1000}=0.08=8\%$$

Interpretation:

Eight percent of customers left during the period.

Why Churn Matters

Suppose a company acquires:

$$100\ customers$$

but loses:

$$100\ customers$$

Net customer growth equals:

$$100-100=0$$

The business is not growing.

Reducing churn is often more profitable than acquiring additional customers.

Statistical Thinking in Marketing

A data scientist asks:

Did sales increase because advertising worked?

Or did sales increase due to random variation?

Suppose average sales before a campaign were:

$$Mean\ Sales=100$$

After the campaign:

$$Mean\ Sales=115$$

The increase appears positive.

However, we do not yet know whether the increase is statistically meaningful.

Future lessons will introduce:

• Hypothesis testing
• Bayesian inference
• A/B testing
• Causal inference

These methods help determine whether observed changes are real.

Example Business Case

A retailer runs an advertising campaign.

Results:

Conversion Rate

$$Conversion\ Rate=\frac{2000}{50000}=4\%$$

CAC

$$CAC=\frac{80000}{2000}=40$$

ROAS

$$ROAS=\frac{400000}{80000}=5$$

Interpretation:

• Cost per customer = $40
• Every advertising dollar generated $5 revenue
• The campaign appears profitable

Healthcare Example

Suppose a hospital launches a screening campaign.

Campaign cost:

$$20,000$$

Patients screened:

$$1,000$$

Cost per screened patient:

$$Cost\ Per\ Patient=\frac{20000}{1000}=20$$

Healthcare organizations use marketing analytics to evaluate:

• Screening effectiveness
• Cost efficiency
• Long-term patient outcomes

Supply Chain Example

Suppose a retailer advertises slow-moving inventory.

Advertising spend:

$$10,000$$

Additional sales generated:

$$60,000$$

ROAS:

$$ROAS=\frac{60000}{10000}=6$$

Interpretation:

Every advertising dollar generated six dollars in revenue while reducing inventory aging.

Python Example

spend = 80000
visitors = 50000
purchases = 2000
revenue = 400000
conversion_rate = purchases / visitors
cac = spend / purchases
roas = revenue / spend
print(f"Conversion Rate: {conversion_rate:.2%}")
print(f"CAC: ${cac:.2f}")
print(f"ROAS: {roas:.2f}")

Output:

Conversion Rate: 4.00%
CAC: $40.00
ROAS: 5.00

Key Takeaways

The most important marketing metrics introduced in this lesson are:

$$CTR=\frac{Clicks}{Impressions}$$

$$Conversion\ Rate=\frac{Purchases}{Visitors}$$

$$CAC=\frac{Marketing\ Spend}{Customers}$$

$$LTV=Average\ Revenue\times Customer\ Lifetime$$

$$ROAS=\frac{Revenue}{Advertising\ Cost}$$

$$Churn=\frac{Lost\ Customers}{Starting\ Customers}$$

A modern Data Scientist is ultimately trying to improve these metrics through:

• Statistical analysis
• Experimentation
• Machine learning
• Optimization
• Business decision-making

Exercises

1. A company spends $120,000 and acquires 400 customers. Calculate CAC.
2. A campaign generates $900,000 in revenue from $150,000 in advertising spend. Calculate ROAS.
3. A subscription company starts with 5,000 customers and loses 250. Calculate churn.
4. If $$LTV=2000$$ and $$CAC=300$$, is the business healthy?
5. Write Python code that calculates Conversion Rate, CAC, and ROAS for any marketing campaign.

References

1. Farris, Paul W., Bendle, Neil T., Pfeifer, Phillip E., Reibstein, David J. Marketing Metrics: The Manager’s Guide to Measuring Marketing Performance. Pearson Education.
2. Kotler, Philip, Keller, Kevin Lane. Marketing Management. Pearson.
3. Gupta, Sunil, Lehmann, Donald R. Managing Customers as Investments: The Strategic Value of Customers in the Long Run. Wharton School Publishing.
4. McCarthy, E. Jerome, Perreault, William D. Basic Marketing: A Marketing Strategy Planning Approach. McGraw-Hill.
5. Montgomery, Douglas C. Design and Analysis of Experiments. Wiley.
6. Gelman, Andrew, Carlin, John B., Stern, Hal S., Dunson, David B., Vehtari, Aki, Rubin, Donald B. Bayesian Data Analysis. CRC Press.

Next Lesson

Lesson 2: Customer Lifetime Value (LTV), Retention Analytics, and Churn Modeling for Data Scientists.