The Pepsi Challenge was one of the most successful advertising campaigns in history.

The Pepsi Challenge was originally invented in Dallas, Texas in 1975 to see if the South truly loved Coca-Cola for its taste, or if it was out of loyalty and good marketing. It turned out that a majority of Americans, not just Southerns, preferred the taste of Pepsi. Coke relentlessly attacked the results, but the findings were indisputable. Worse still, it turned out a majority of ‘Coke Drinkers’ preferred the taste of Pepsi. Unable to blunt Pepsi’s momentum, Coke decided to focus instead on improving their own product, and so launched New Coke.

This Case Study provides a high-level overview of the workflow without detailed explanation. It assumes you are already somewhat familiar with KNIME and Market Simulation. If not, start by reviewing the Building Blocks and Community Nodes.

According to Wikipedia:

The 1975 Pepsi Challenge took the form of a Single Blind Taste Test. At malls, shopping centers, and other public locations, a Pepsi representative would set up a table with two white cups: one containing Pepsi and one with Coca-Cola. Shoppers were encouraged to taste both colas, and then select which drink they prefer. Then the representative reveals the two bottles so the taster can see whether they preferred Coke or Pepsi. The results of the test leaned toward a consensus that Pepsi was preferred by more Americans.

This Market Simulation is built upon Coke and Pepsi data from the time of the original Pepsi Challenge. In 1975 the Price of a 6-pack of 12oz Pepsi cans was 88 cents. At that time also, more Customers preferred Coke over Pepsi, so in this model Coke has a higher average WTP.

The results from the Simulate Market node reflect the 1975 Market Share in the USA when Coke vs Pepsi sales were about 60:40.

1975 Results

Before the Pepsi Challenge, Coke had a 60% share of the Market.

The Pepsi Challenge asked Customers to pick the cola they preferred. The results from the Taste Test found that Coke Drinkers only guessed correctly about 47% of the time. In contrast, Pepsi Drinkers guessed correctly about 86% of the time. These guesses were replicated in the KNIME ‘Conditional Label Assigner’ node.

The ‘Rule Engine’ then evaluates whether the Customer guessed correctly and adds a Correct column. The ‘Row Splitter’ then splits those Customers who guessed incorrectly to the bottom port.

The logical heart of this Market Simulation is located within the Scale Purchased node. If the Customer guessed incorrectly then increase the Customer’s WTP for their preferred cola (ie the cola they guessed) by 10%. If the Customer guessed correctly then their WTP for both colas is not changed.

The final results from the Market Simulation replicate the real-world results, with Pepsi gaining about 6% Market Share against Coke.

Final Results

Pepsi raised their Market Share by about 6% vs Coke.