COLLECTING AND ANALYZING DATA|
Data Mining for the Rest of Us
Mary has a little coffee shop. She sells coffee, of course, in many interesting variations, as well as bagels, muffins, doughnuts and similar fare geared toward the morning and early afternoon appetite.
Her shop is in a downtown location, with many nearby workplaces. Her hours of operation are 6 am until 4 pm Monday through Friday.
Although Mary felt she knew her customers pretty well, she decided to take some time to track what she knew (or thought she knew) about these customers — who buys what, when they buy, what they spend — to and see if she could design some marketing efforts to get more folks to buy more of her products more often.
Collecting and analyzing customer data can be extremely useful in identifying (or confirming) buying habits so that marketing efforts can be better targeted to achieve specific goals.
Based on her general customer knowledge, Mary designed the data collection to add more detailed information around what she already knew; for example, more sales come earlier in the day, customers are more made up of men, coffee is more popular than tea, and bagels are her best seller. What she wants is the data to confirm these assumptions, in a way that allows her to create highly targeted promotions around the specifics.
By collecting the additional specifics, Mary was able to then "mine" the detailed data and decide what she could do to boost sales?
First, for the low hanging fruit, notice that the Time of Purchase column trails off by dollar as the day goes on. Mary observed that fewer of the folks coming in late in the day buy a food item, and she always has left-over food (and more on Mondays and Tuesdays). Mary decides she should begin by giving the 12% of folks who come in from 2-4 pm an incentive to buy food. She institutes a bold, 50% off sale on all food purchased with a beverage after 2 pm.
Second, Mary was surprised that a full 60% of her customers buy only a beverage. She decided to institute a card/sticker rewards program, whereby a person earns a sticker with every food/beverage purchase, and with five stickers the customer earns the free beverage of her choice.
Third, Mary decides to try a limited promotion related to tea purchasers. She creates a "tea for two" offer that includes two teas and two food items of the customer's choice, for $4.99 on Fridays. While this represents about a 40% discount, she postulates she may be able to bring in more tea customers. Since tea has a lower cost of goods sold than coffee, her profit reduction on the combination is significantly lower than the 40% cost.
She promotes the offering with a flyer handed out in store on Fridays for four weeks and does 2 unaddressed mail drops to all businesses and residences within 5km of her shop location.
Data collection and mining need not be complicated. Mary's simple data collection effort helped her confirm some facts she thought she knew — most customers were men, more sales come earlier in the day, and coffee was more popular than tea. The data collection also taught her a few key things she wasn't aware of — more than 60% of customers only bought a beverage and certain days would yield more left-over food product. Combining all of the information she now had, Mary was able to identify areas for improvement and formulate some great ideas for addressing the specific issues and improving sales.
What do you think? Feel free to contact us with data questions and/or marketing suggestions for Mary.
Mary printed out a simple data tracking worksheet where she could note the following:
The form is designed to provide Mary with lots of data, while making it easy and fast to fill out as part of the sales process. That's why she used checkboxes for the various notations. It's faster than writing the answers by hand.
Mary collected information from 1,274 customers over a period of seven days. Here are some highlights:
GENDER OF CUSTOMER:
AGE OF CUSTOMER:
PURCHASE TYPE/ BEVERAGE BUYERS:
PURCHASE TYPE/ FOOD BUYERS:
TIME OF PURCHASE/ BY DOLLAR:
TOTAL PURCHASE AMOUNT: