DISCOUNTING INTENT WITH SEGMENTATION $-$ AN EXCEL PRICING EXAMPLE

Many experiments measure intent as a business outcome because it is easier to ask potential customers about their intent to purchase in a survey than to measure an actual purchase. For example, we might ask customers their intent to purchase a product on a scale from $1$ to $7,$ where $1=$very unlikely and $7=$very likely. Oftentimes, this is analyzed without discounting intent. However, it is better to apply a discount because people do not do what they say they will do$.$ In other words, if someone responds with a $7$ out of $7,$ yes, they are very likely to purchase, how many of them would actually purchase?

The answer is usually less than $100\%.$ Consider ACNielsen BASES Predicted Purchase Intent: $7$ Very Likely to Buy Product $0\mathrm{.}756$ Likely to Buy Product $0\mathrm{.}505$ Somewhat Likely to Buy Product $0\mathrm{.}254$ Undecided $0\mathrm{.}13$ Somewhat Unlikely to Buy Product $0\mathrm{.}052$ Unlikely to Buy Product $0\mathrm{.}021$ Very Unlikely to Buy Product $0\mathrm{.}01$ This means that we expect only $3$ out of $4$ people who say they are $$very likely$$ to actually make the purchase when given the real chance. Analytically we must adjust the outcome to reflect this. Use the data collected on intent in the Assignment file $$Assignment #$3$ Intent Excel Sheet$$ and adjust for discounting intent as follows:

$1.$ For each price point: a$.$ Apply $$discounting intent$$ adjustments to predict estimated conversion

b$.$ Estimate promotion expenditures: How much money will you spend on promotion? A good # for this example can be $$15,000$

c$.$ Estimate estimated exposure: How many people representative of your experiment$$s sample will see the ad based on promotion expenditures? Let's use $1\mathrm{.}5$M d$.$ Estimate COGs $($variable & fixed costs$)-$ let's use $$40$ for this example e$.$ Calculate profit $=($price $$ COGs $/$ costs$)*($estimated quantity$)$ fixed costs

$2.$ You should now have a calculated gain for marketing to both segments. Note this total $ figure.

$3.$ Now calculate separately for those in segment $1$ versus those in segment $2($just copy the spreadsheet, sort the copy by segment and retain only the records of a segment, deleting the others$).$ Compare the total $ gains with a chart; what segment$($s$)$ should you target? Note: The Intent Excel Sheet is a file with results from an experiment in which $130$ people were exposed to a promotional message for the same product but with two possible price points, $$45$ and $$65.$ People were randomly assigned to see EITHER $$45$ or $$65$ as the price and then responded to the question, $$How likely are you to purchase this product?$$ using a $1$ to $7$ scare, with $1=$very unlikely and $7=$very likely scale. People were identified as being in one of two segments. DISCOUNTING INTENT WITH SEGMENTATION $-$ AN EXCEL PRICING EXAMPLE

Many experiments measure intent as a business outcome because it is easier to ask potential customers about their intent to purchase in a survey than to measure an actual purchase. For example, we might ask customers their intent to purchase a product on a scale from $1$ to $7,$ where $\backslash (1=\backslash )$ very unlikely and $\backslash (7=\backslash )$ very likely.

Oftentimes, this is analyzed without discounting intent. However, it is better to apply a discount because people do not do what they say they will do$.$ In other words, if someone responds with a $7$ out of $7,$ yes, they are very likely to purchase, how many of them would actually purchase? The answer is usually less than $\backslash (100\backslash \%\backslash ).$ Consider ACNielsen BASES Predicted Purchase Intent:

$-7$ Very Likely to Buy Product $0\mathrm{.}75$

$-6$ Likely to Buy Product $0\mathrm{.}50$

$-5$ Somewhat Likely to Buy Product $0\mathrm{.}25$

$-4$ Undecided $0\mathrm{.}1$

$-3$ Somewhat Unlikely to Buy Product $0\mathrm{.}05$

$-2$ Unlikely to Buy Product $0\mathrm{.}02$

$-1$ Very Unlikely to Buy Product $0\mathrm{.}01$

This means that we expect only $3$ out of $4$ people who say they are "very likely" to actually make the purchase when given the real chance.

Analytically we must adjust the outcome to reflect this. Use the data collected on intent in the Assignment file "Assignment $\backslash$#$3$ Intent Excel Sheet" and adjust for discounting intent as follows:

$1.$ For each price point:

a$.$ Apply "discounting intent" adjustments to predict estimated conversion

b$.$ Estimate promotion expenditures: How much money will you spend on promotion? A good $\backslash$# for this example can be $\backslash (\backslash$$ $15,000\backslash )$

c$.$ Estimate estimated exposure: How many people representative of your experiment's sample will see the ad based on promotion expenditures? Let's use $1\mathrm{.}5$ M

d$.$ Estimate COGs $($variable $\backslash$& fixed costs$)-$ let's use $\backslash (\backslash$$ $40\backslash )$ for this example

e$.$ Calculate profit $\backslash (=(\backslash$text $\{$ price $\}-\backslash$text $\{$ COGs $\}/\backslash$text $\{$ costs $\})^\{*\}(\backslash )$ estimated quantity $\backslash ()-\backslash )$ fixed costs

$2.$ You should now have a calculated gain for marketing to both segments. Note this total $\backslash (\backslash$$ $\backslash )$ figure.

$3.$ Now calculate separately for those in segment $1$ versus those in segment $2($just copy the spreadsheet, sort the copy by segment and retain only the records of a segment, deleting the others$).$ Compare the total $\backslash (\backslash$$ $\backslash )$ gains with a chart; what segment$($s$)$ should yo