Tuscan Lifestyles: Targeting Customers with RFM Analysis

Part I: Preliminary and Decile Analysis

1. What percent of customers responded (i.e. bought anything) from this catalog?

2.5% of the customers bought from this last catalog.

Analyze → Descriptive Statistics → Frequencies

2. Of those who bought, what was the average dollars ordered from this catalog? The average dollars ordered from this catalog was $104.24 per buying customer.

Analyze → Reports → Case Summaries

3. Create decile variables for recency, frequency and monetary.

Transform → Rank Cases

4. Create a bar chart showing the response rate to this catalog by recencydecile.

Data → Aggregate

Graph → Legacy Dialogs → Bar

5. Create a bar chart showing the response rate to this catalog by frequency decile.

Data → Aggregate
Graph → Legacy Dialogs → Bar

6. Create a bar chart showing the response rate to this catalog by monetary decile.

Data → Aggregate

Graph → Legacy Dialogs → Bar

7. Using only those customers who placed an order from this catalog, create bar charts showing the average dollars ordered from this catalog by recency, frequency andmonetarydeciles.

Data → Select Cases

Graph → Legacy Dialogs → Bar

8. What do the above bar charts reveal about the likelihood of response and the size of the order across the different recency, frequency, and monetary deciles?

Part II: RFM Classification

9. Generate a report showing the number of customers, the number of buyers, and the response rate for each RFM cell (using the sequential n-tiles approach as recorded in the RFM_SEQ variable).

Analyze → Reports → Case Summaries

Part III: Profitability Analysis

Now, compare the profitability of two scenarios:
• mailing to all 96,551 customers
• mailing only to those customers in RFM cells (using the sequential n-tiles approach) with a response rate that is equal to or greater than the breakeven response rate

10. Use the following costs to determine:
(a) the gross profit in dollars,
(b) the gross profit as a % of gross sales, and
(c) the return on marketing expenditures (gross profit/cost to mail catalogs) as a result of mailing the catalog to all 96,551 customers:
Cost to produce and mail catalog = $1
COGS and variable costs on orders = 50% (of sales revenues)

11. What is the breakeven response rate?
Breakeven response rate = cost to mail catalog / net profit of the average sale
$1 / ($104.24 * 50%) = 1 / 52.12 = 0.0192

12. Compute the following items. You can do these (1) by hand (not recommended), (2) in Excel (using the results from question 9 above) or (3) you can create an aggregate SPSS dataset. If you want SPSS to do the bulk of the calculations, follow the instructions in Exhibit 3 - and then you can use analyze ... reports... case summaries for the final computations:

• Determine which RFM segments (using the sequential n-tiles approach) have response rates exceeding the breakeven rate.
• Determine the number of customers belonging to these profitable segments.
• Determine the number of buyers belonging to these profitable segments.
• Finally, what would the

(a) the gross profit in dollars,
(b) the gross profit as a % of gross sales, and
(c) the return on marketing expenditures (gross profit/cost to mail catalogs) have been as a result of mailing the catalog only to those customers in the RFM cells with response rates exceeding the breakeven? That is, rather than mailing to all 96,551 customers - what would the profitability of the mailing have been if mailed to the subset of customers in ‘profitable' segments?

13. Examine the first 20 or so observations in the database.

What do you notice about theRFM1 and RFM2 values?