Classification used in Marketing

Goal: Utilize KMeans Clustering, RandomForest, and other Classification methods to solve marketing problems such as Market segmentation and predict Competitive marketing spend

Data Collection

• Top 30 tech companies' info from Fortune.com
• Competitive Advertising Spend from research firm Kantar
• TV/Radio Audience Population from research firm Nielsen

Data Prepration

Data Exploration

Media Mix and Market Mix of top 30 tech companies in 2019

While some companies spent higher % in National advertising, some other companies focused on local marketing strategies

Part I: Build a classification model to predict national spend range for top tech companies

Notebook link

Part II: Clustering 210 US Local DMAs to several segments

I am trying to create clusters for US 210 DMAs based on tech industry advertisers' local spend to have a deeper understanding on the similarities among the markets.

The implication can be: find out market groups that tech advertiser prioritize or heavily advertise in; create local marketing strategies that cater to each group.

Notebook link