Marketing examples

Data-Driven Decision Making for Business

Ted Kwartler

Data Dude

4 P's of marketing

1) Promotion

2) Pricing

3) Product

4) Placement

To grow or not to grow

• Should we grow at great expense?
• Should we focus on profitability?

Modeling customer behavior

Customer propensity model: provide a likelihood a potential customer will accept the offer

Modeling customer behavior

Customer propensity model: provide a likelihood a potential customer will accept the offer

Understanding a propensity function

$$

• Probabilities are between 0 and 1
• Most customers don't click on the ad

Identify the tail, most likely responses

$$

• Probabilities are between 0 and 1
• Most customers don't click on the ad
• If 95% of the time customers don't click, an accurate model would predict no-clicks
• So focus on the accuracy in the top decile (blue box)

Model predictions by expected responses

Return on investment

How much money is returned for each spent dollar?

Costs and revenue

Let's compare cost and revenue

Profit curve

Ultimately, the model propensities by the profit are important

Off to it!

Data-Driven Decision Making for Business