Analysis of Markets

You are interested in examining the effect of in-store displays on sales and have the following results from the model: ysales=β0+β1xdisplay and receive the following: Constant: 1358 Display: 178 Now imagine that you reran this same model and analysis over a smaller sample of 100 stores from the same 52 weeks and found that the estimate for the display coefficient is exactly the same (β=178). Based on this information and the assumption that the two samples were drawn from the same population, which of the following statements about the results from that second model is LEAST LIKELY to be true. The standard error for the display coefficient in the second model is larger than the standard error for the display coefficient in the first model The t statistic for the display coefficient in the second model is smaller than the t statistic for the display coefficient from the first model The p-value for the display coefficient in the second model is larger than the p-value for the display coefficient from the first model The 95% confidence interval for the display coefficient in the second model is narrower than the 95% confidence interval for the display coefficient in the first model

The 95% confidence interval for the display coefficient in the second model is narrower than the 95% confidence interval for the display coefficient in the first model

All of the following are true about the concept of revenue management except... Select the best answer... Revenue management is quite important in sectors with high fixed costs where asset utilization is paramount Revenue management principles emphasize a segmented approach to setting prices Revenue management emphasizes a dynamic approach to pricing optimization The goal of revenue management is to find the right price that the customer will pay and charge everyone that same amount

The goal of revenue management is to find the right price that the customer will pay and charge everyone that same amount

If you have a sample of 200 individuals and had asked them their willingness to recommend your brand to a friend or colleague on a scale from 0 to 10, all of the following are true except... Passives would be those who answered 7 or 8 to the question The net promoters score would be the number of respondents who answered 9 and 10 subtracted from the number of respondents who answered 0 through 6 Detractors would be anyone who answered less than 7 to the question It is possible to get a negative net promoters score using this type of data

The net promoters score would be the number of respondents who answered 9 and 10 subtracted from the number of respondents who answered 0 through 6

A company pays $40 per unit to acquire a product from a wholesaler and charges its own customers a price of $50 per unit for the product. Assuming these are all the facts you know, which of these statements is true about this situation? The percentage markup for this product is larger than the percentage margin for this product The unit margin is $40 The percentage markup is 20% The percentage markup for this product is the same as the percentage margin for this product

The percentage markup for this product is larger than the percentage margin for this product

A paid search ad was displayed 2,000 times, clicked on 75 times, which lead to 0 completed sales. This activity generated $15 in click charges. What is the cost per click for this ad? $0.20 $0.02 $0.008 $0

$0.20

Assume you have a product priced at $4 with a 9% profit margin (contribution margin). You plan to cut the price of this product by $0.20. On average you have historically sold 10,000 of these items per week. Without any changes to average weekly volume, what is the dollar change in weekly contribution (revenue-variable costs) due to this proposed pricing change? $2,000 $1,600 $900 $180

$2000

Assume you have the following weekly sales figures: [$1000, $4000, $5000, $7000, $3000] What is the average weekly sales revenue over this five week period? $1000 $4000 $5000 $6000

$4000

The current year's sales are $3,000,000 and 20% growth is anticipated for next year. Given this anticipated growth, how much (in dollars) is sales expected to increase next year? $100,000 $500,000 $600,000 $2,500,000

$600,000

A paid search ad was displayed 1,000 times, clicked on 100 times, which lead to 1 completed sale. This activity generated $10 in click charges. What is the click-through-rate for this ad? .1% 1% 10% 100%

10%

If a company spent $20,000,000 in fixed cost on a product with a $0.40 unit margin and has average daily sales of 500,000 units for that product, how many days to break-even on this fixed cost expenditure? 1 10 100 1000

100

B/E (100k upfront)

100/WAUM

You run a regression model and get the following output. Degrees of Freedom Sum of Squares Explained 3 xxxxxxxxxx 300 Unexplained 997 xxxxxxx 1200 In this model, how much variance in the dependent variable is explained by the model? 3% 20% 25% not enough information

20%

You estimate the following regression... yunitsales=β0+β1 xprice + β2 xfeature + β3xdisplay and receive this output: xxxxxxxxx coeff constant xx 1100 price xxx -200 feature xx -11 display xx 185 All else constant, at a price of $4 and no feature nor display what is the predicted amount of unit sales?

300 units

In thinking about using the chain ratio for opportunity assessment, if the US market is 330 million persons and the product is expected to target persons under age 25 (25% of the population), with less than the median income, who don't already own an iPhone (65% of the population) which formula will give the estimated market size for the product using the stated information? 330MM x .75 X .50 X .35 330MM x .75 X .35 330MM x .25 X .50 X .65 330MM x .25 X .65

330MM x .25 X .50 X .65

Imagine you have this table. Assuming there are 1,000 records in this dataset how many records in the data have a Feature value equal to Yes? xxxxxxxxxxDisplay xxxxxxxxxNoxxYes FeatureNo 47% | 15% 62% FeatureYes 20% | 18% 38% xxxxxxxxxx67% | 33% 100% Select the best answer... 180 374 380 38%

380

Consider the following linear demand function: Q=β0+β1price . You are given price and quantity data and have the resulting parameter estimates from a linear model: Q=1575−175×price . Given this information, what is the price that maximizes revenue? (you do not need a dollar sign in your answer)

4.5

Baseline sales for a particular week is estimated to be 2,000 units. You actually sell 2,400 units that week. How many incremental units did you sell? 20% 25% 400 1200

400

Assume you have a product priced at $600 with a 25% profit margin (contribution margin). Your variable costs are set to increase by 10%. How much more would you need to sell to break even relative to the current scenario for this increase in variable costs? 10% 43% 67% 70%

43%

Given the following output from a logistic regression, what percent of observations are correctly classified by this model? xxxxxxxxxActual xxxxxxxxxxxxxx0xxxx1 Predicted 0xx60xxxx40 Predicted 1xx30xxxx70 70% 65% 60% 35%

65%

These three stores comprise a market (total $ sales): A ($3MM), B ($6MM) and C ($1MM). Your product is sold in stores B & C. What is your product's %ACV? $10MM 70% 90% 60%

70%

If a company has an existing product with a $2 unit margin and launches a new product in the same category with a $1.50 unit margin, what is the break-even cannibalization rate? 133% 25% 50% 75%

75%

Imagine you have a vector of binary data in Excel. All of the following are possible methods to determine how many of the values in that data vector are equal to 1 using just the listed function, EXCEPT... =COUNTIF() Pivot Table =SUMIF() =AVERAGE()

=AVERAGE()

Time to Cover

BE/Sales Velocity

Market Share

Bounty/All Paper Towels

Increase in cost up 2%

Change VC x 1.02% Price-New VC = New Margin

All of the following are valid concerns about correlation (r) except... Correlation can be induced by the presence of outliers Correlation can only capture positive linear associations Correlation can only capture linear associations Correlation can be suppressed by the presence of outliers

Correlation can only capture positive linear associations

All of the statements below are true EXCEPT... Data is information with context A variable is an object denoting an event, act, characteristic or quantity that can be measured and assigned a value The analytics triad includes acquisition, analysis and visualization General types of data include image, text and numeric

Data is information with context

You have some data for an A/B test where the dependent variable is clicks. You run a linear regression model where the dependent variable is clicks and the independent variable is the version (A=0 and B=1) and get the following output. Intercept 52.66 Version -2.52 If, instead, you re-ran the linear regression model after making changes to the independent variable Version such that Version is equal to zero if B and Version is equal to 1 if A, which if the following statements is FALSE? The estimate of the Version parameter will be the same, only the sign will change The estimate for the intercept will be the same The t statistic for the Version paramter will be the same, only the sign will change The p-value for the Version parameter will be the same

Estimate will be the same

Two firms compete with each other selling normal goods in a market vertically differentiated on price. Firm A sells 1000 units at a price of $500 and Firm B sells 500 units at a price of $1000. The unit volume-based market share for Firm A is lower than the revenue-based market share for Firm B. TRUE/FALSE

FALSE

In a typical two-dimensional dataset, the rows refer to the variables and the columns refer to the unique observational units. TRUE FALSE

False

A retail marketing activity that includes ads for products often appearing as inserts in print media is a... Feature Display TPR Coupon

Feature

If you estimate a sales response model on weekly store-level data that has an overall intercept as well as store-specific intercepts for all but one (baseline) store in the dataset, this is an example of... Select the best answer... a latent-class model a fixed-effects model a random-effects model a discrete-choice model

Fixed-effects model

All of the following are true about residuals except... Homoscedastic residuals indicate a model passes at least one econometric assumption of the linear model For a given model specification, the sum of squares of the residuals is maximized at the estimated regression line If the model estimated parameters are known, it is possible to compute residuals from the data Residual = observed value - fitted value

For a given model specification, the sum of squares of the residuals is maximized at the estimated regression line

All of the following statements are true, except... One way that paid search advertising differs from the traditional advertising model is that paid search ad impressions can be more directly linked to online purchase transactions. For keywords on Google, the rank of a paid search keyword ad is a combination of the keyword bid and the quality score. For keywords on Google, the rank of an organic search keyword is a combination of the keyword bid and the quality score. One way that paid search advertising differs from the traditional advertising model is that the advertiser pays for clicks rather than impressions.

For keywords on Google, the rank of an organic search keyword is a combination of the keyword bid and the quality score.

If you knew that your buyer index for Hispanic/Latinos was 160 and also knew that the base rate of Hispanic/Latinos in the market was 20%, then the proportion of your buyers who are Hispanic/Latino is 32%. TRUE FALSE

TRUE

Product A sold 9,000 8 ounce packages for $2.50 each. Product B sold 6,000 16 ounce packages for $4 each. In a product category where 16 ounces is the equivalent unit, Product A has a higher price per equivalent unit. TRUE FALSE

TRUE

Starting with a basic regression model: yunitsales=β0+β1xprice , if there is an additional independent variable (x) that is currently not in the model, and this x variable is positively correlated with your dependent variable (y), and negatively correlated with the other independent variable, price, this is an example of omitted variable bias. TRUE FALSE

TRUE

How many GRP's does a television advertisement that appeared 10 times on a television show that reaches 16% of the total United States audience deliver? 16 GRP's 160% GRP's 160 GRP's Would need to know what percentage of that audience is the target market first

Tbd -- Ty should be emailing me back

When talking about "big data", although some of the aspects present technological challenges to overcome, according to our discussion in class, which aspect of "big data" presents the most interesting potential for developing new marketing insights? Volume Velocity Variety Veracity

Variety

You estimate the following regression model.. ln(yunitsales)=β0+β1ln(x(own price)) + β2 ln(x(competitor price) + β3xdisplay and get the following output.. Constant 8.0 own price -2.0 comp price .5 display 0.4 When you have a display what affect does it have? You sell 40% more units You sell 49% more units You sell 4,447 units There is no effect on unit sales

You sell 49% more units

With regard to different data collection techniques, a mail survey is an example of ____ and clickstream (web browsing) data is an example of ______ . monitoring ; communicating communicating ; monitoring discrete data ; continous data unstructured data ; structured data

communicating ; monitoring

In the following model: ln(yunitsales)=β0+β1ln(price) The coefficient β1 is best described as.. slope elasticity

elasticity

You have weekly store-level unit sales, price and advertising data and estimate the model: ln(unitsalesstore,week)=β0+β1ln(price)+β2ln(GRP) and get the following results... Intercept: 12.2 Ln(price): -1.5 Ln(GRP): xx 0.2 Relative to current levels of 200 GRPs and price of $106, if the level of advertising were increased by 10% what can be said about the result? increasing advertising support by 10% would result in a 20% increase in sales increasing advertising support by 10% would result in an average per-store increase in unit sales of 10 units increasing advertising support by 10% would result in an average per-store unit sales of 525 units increasing advertising support by 10% would have no apparent effect on unit sales

increasing advertising support by 10% would result in an average per-store increase in unit sales of 10 units

In the basic regression model: ysales=β0+β1xprice all of the following statements about β1 are true EXCEPT... it is the the intercept of the linear function it is the slope of the linear function it is the rate of change in y for a unit change in x it is the change in unit sales for a dollar change in price

it is the intercept of the linear function

You estimate the following regression model... ln(yunitsales)=β0+β1ln(x(own price)) + β2 ln(x(competitor price) + β3xdisplay and get the following output: xxxxxxx coeffxxxx std constant 8.0 xxxx 0.50 own price -2.0 xxx 0.20 comp xxx 0.5 xxxx0.05 display xx0.4 xxxx 0.04 If your own price is $5 and your competitor's price is $5, if there is no display, what is the model predicted unit sales?

loss of 267 units

A database in which two or more tables are linked together on key variables is a _____ database.

relational

A Q-Q plot of the residual values is most useful for diagnosing which econometric assumption of an estimated linear model? Normality Homoskedasticity Independence Linearity

Normality

Your company is considering two different new product launches in a category where you already have existing products. Potential new product A has a 20% margin, while potential new product B has a 30% margin. If both of these products were compared to the same existing product data, which potential new product would have the higher break-even cannibalization rate? Potential New Product A Potential New Product B Not Enough Information

Not Enough Information

Variable Cost

P- Unit Margin

Unit Margin

Price x Profit Margin

Using household panel data, you observe two (of many) households (A and B) that both start purchasing in the product category in the exact same week. Household A returns to the market every three weeks for the next 27 weeks and on each purchase occasion buys 1 unit of your brand. Household B returns to the market every 9 weeks for the next 27 weeks and buys 3 units of your brand on each purchase occasion. Which of the following statements is most certainly true? Purchase frequency for household A is lower than purchase frequency for household B Total brand volume for household A is greater than total brand volume for household B Purchase size for household A is smaller than purchase size for household B Penetration for this product category is 100% Purchase size for household A is smaller than purchase size for household B

Purchase size for household A is smaller than purchase size for household B

All of the following web traffic metrics are most likely positively correlated with a high bounce rate EXCEPT... (select the best answer) Low average time on site Low average pages per session Low amount of scrolling on the page Low number of unique visitors to the page

Low number of unique visitors to the page