Problem Description A firm wants to analyze the impact of its social media efforts on consumer behaviors. The firm starts its social media presence by initiating a Facebook page. Some of its consumers become fan of the firm's Facebook page (termed as treatment group) while others do not (termed as control group). Note that consumers in both these groups continue to purchase from the firm before and after the initiation of its social media presence. The firm collects the following data to address this questions. Data ConsumerID Group Period Week Spending 10001 27.88 10001 25.96 10001 51.62 10001 29.56 10001 10 59.76 10001 11 19.71 10001 14 52.30 10001 16 34.42 10001 17 48.26 10001 19 39.88 10001 22 85.63 10001 25 33.26Data Description A marketing manager from the rm has selected scanner panel data from two-year period for this analysis. In the rst year [1-54 weeks], rm did not have social media presence while for the next year [55-103 weeks)1 rm has its social media presence. Thus, assuming the event, consumer participation in social media, takes place in week 55 for all the consumers in the treatment groups. The manager randomly selects 200 consumers from each group and observes their purchase behaviors. The description of the variable are as follows: . ConsumerID: Identication code for consumer in Group: 1, if Cmsnmerf E TreetmentGronp. Gran = p {0, otherwise. . Period 1, if Week > 55. Period = 0, otherwise. . Week: Purchase occasion of consumers at weekly level. . Spending: Dollar spending by individual consumers during each purchase occasion. Task Analyze whether the firm's social media effort is a good marketing strategy or not. Answer this question in terms of changes observed in behaviors of the panel of consumers captured by the marketing manager. . Define metrics how you would capture the success of social media marketing with the given data set. . This is a quasi-experimental design. Perform difference in difference analysis. Provide table similar to the following: Difference in Difference Table After Change Before Change Difference Group 1 Yl1 Y10 AY = YlI - Y10 (Treat) Group 2 You You AYe = Yo1 - You (Control) Difference AAY = Yt - YeDeliverables This is individual work. Students will submit the following deliverables for this assignment: - A brief report containing the following: o Briey describe the data (Max 300 words} {20 points) 0 Analyze the data [present the table} {50 pointsi o Conclusion {Max 500 words} [30 points) Note that there is no need to show your raw analysis [such as coder steps involved in analysis etc.} Your conclusion should include insights into the following: - Insights from the data analysis? - 1il'luihy such an analysis is important to capture the ROI of social media marketing? - Thoughts into any other optimal way to measure the ROI of social media marketing. Econometric Theory Impact evaluation of any policy intervention is one of the primary interests of policymakers- Such an analysis leads to l evidencebased pol icy making I. In marketing, managers are constantly overwhelmed with various types of decision-making such as campaign design, new product development, market expansion, and marketing promotions. One of the key questions that marketing managers need to know is "What is the impact of their decision-making on the firm's bottom line?\" In marketing, practitioners have widely used AgB testing 3 to compare the two versions of some decision- making le.g., whether to have a white vs. blue background for the website} to determine which has better performance. MB testing is the basic kind of randomized controlled trial {RCTs}. The basic idea of RCT design is as follows: 1. Treatment: Carefully formulate the intervention le.g., whether to give price promotion or not?} 2. Randomization: Randomly split the population into two groups: :- Treatment group: Those who receive the treatment [e.g., get a promotion}. 0 Control group: Those who don't receive the treatment [e.g., do not get a promotion} 3. Measure the outcome of interest for both the groups, i.e., control and treatment groups, :- before the treatment has taken place :- after the treatment has taken place 4. Compare the outcome across periods and across groups. Difference in differences {DID} is one of the methods to carry out the last step of comparison