(a) What is the main finding of this study? Why is measuring the effects of advertising difficult (make sure to read and understand the discussions following equation (7))? Why is it easier to measure the effects of online advertising than the effects of traditional TV advertising? (15 points)

(b) Suppose you are a consultant who helps CPG (consumer-packaged-goods) firm run advertising experiments. You are currently working with two firms.

Firm 1 plans to run their A/B experiment on 1000 customers. From prior experiments, you expect the average value of purchases for consumers in the treatment group (receive ad) to be

\$0.30, while the average value of purchases for consumers in the control group (no ad) to be

\$0.20. Assume that the standard deviation of the average purchase value for both treatment and control groups is \$5.00.

Firm 2’s plans to run their A/B experiment 10000 customers. Firm 2’s product is more expensive but purchased less frequently than Firm 1’s product. Coincidentally, the average value of a purchase in the treatment groups (receive ad) and control groups is the same as it is for Firm 1. However, the standard deviation of the average purchase value for both treatment and control groups is \$25.00.

Which firm’s ad campaign is more likely to show a positive effect? Should either firm expect to see a significant positive effect? (10 points)