A travel services firm has a paid search campaign. Among the many keywords in its campaign, we have data on four keywords, denoted by kw8322228, kw8322392, kw8322393 and kw8322445. These are generic, non-branded keywords, where the prospect's query does not indicate that he/she is leaning toward a specific brand. For each keyword, the firm tried several bid values and recorded the corresponding number of clicks that it received.
The files are named clicksdata.kw8322228.csv, clicksdata.kw8322392.csv, clicksdata.kw8322393.csv, clicksdata.kw8322445.csv respectively and be downloaded here: clicksdata.kw8322228.csv, clicksdata.kw8322392.csv, clicksdata.kw8322393.csv, clicksdata.kw8322445.csv or all together in this zip file: clicksdata.kw.all.zip.
Part A: Estimate the alpha and beta parameters for each of these four keywords for this firm. Hand-in: The eight numbers. No additional writeup required. For instructions on how to estimate alpha and beta from empirical data on bid values and the number of clicks, as we did in class, click here: Estimating alpha and beta.
Part B1: Assume that the LTV dollar value and the conversion rate values for each of the keywords for this firm are as shown in the table below. These numbers are available in an Excel spreadsheet here: hw-kw-ltv-conv.rate-data.xlsx keyword ltv conv.rate kw8322228 354 0.3 kw8322392 181 0.32 kw8322393 283 0.3 kw8322445 107 0.3 Assume that you have no budget constraint. Using the alpha, beta parameters from Part A and the LTV and conversion rate values, estimate the optimal bids for each of the four keywords. Hand-in: the optimal bid value, the corresponding profit and the corresponding total expenditure for each of the four keywords. No additional writeup required.
Part B2: Looking across the four keywords, there is a relationship between LTV and alpha, a relationship between LTV and beta, and a relationship between LTV and the optimal bid. What are these relationships? What are the likely reasons for this relationship? Hand-in: Your identification of the nature of these relationships and your likely reasons. Please do not spend more than 10 minutes on this part. The relationship is easy to spot but the explanation is much less obvious. If one cannot propose the explanation in under 10 minutes, it is unlikely to happen by spending more time on this. This question is on marketing and consumer psychology rather that statistics. Hint: it has to do with consumer segments and the fact that these are generic, non-branded keywords.
Part C: Assume now that you have a budget constraint of $3000 across these four keywords. Compute the optimal bid amounts and the corresponding expenditures for the keywords. Which keyword has the least percentage reduction in expenditure as a result of the constraint? What is the likely reason for this?