Business Intelligence (or Business Analytics) is the process of analyzing data generated in the process of running a business. That analysis will, hopefully, result in information (“information” is data that is useful) that we can use to make better business decisions.
I have a Seller’s Account on amazon.com, and over the past 3 years, I have sold 167 items. [Note for any would-be Amazon Entrepreneurs – the price you sell at, plus the shipping & handling charge, minus Amazon’s cut, minus the actual cost of shipping comes out to about zero. The reason I do this is to get “my stuff” out of my house and into the hands of someone who actually wants it.] This is a sample of the information associated with my sales:
Since I believe that music collections are going to be digital (iTunes on your computer/tablet/cloud/phone, or wma files ripped to your Windows machine), the days of physical cds are soon to be over. [A personal note: when everyone had turntables and records in the late 60’s and early 70’s, I had a tape deck – I would buy an album for $3.25, record it, and then sell it to fellow students for $3.00 – a good deal for them, and a good deal for me. I could get 10 albums on a reel of tape, and no albums to lug around.] As a result, I have many cds for sale on amazon.com, and would like to know when is the best time of day to go into my account and adjust them to be the Low Price (because this is the way I purchase: if the quality is ok, I go for the Low Price).
Being in Massachusetts, I thought that a good time to adjust prices would be either at noon (for the lunch crowd) or 6 PM (for the evening buying crowd). You can see that for each order, Amazon gives me the order date and order time (in Pacific time), along with the item details. For this project I am only interested in the time; additional analysis can certainly be done on day-of-the-week, as well as music cd-versus-book. Maybe I will throw all that data into Tableau for another blog.
I made two Excel spreadsheets. The first spreadsheet had 24 rows (one for each hour), 1 column specifying the hours, 12 columns filled-in for each Amazon page (15 entries per page), and 1 column summing the 12 detail columns. The data in the second spreadsheet is a direct link to the summary column in the first spreadsheet, and then I made a horizontal bar chart for each hour (now in Eastern Time because that is what I understand). As I populated each cell in the first spreadsheet, the bars grew on the right:
When all was done, the 167 orders had two peaks (Noon, and 5 PM), with secondary peaks at 8 PM, 1 PM and 11 PM. Since Amazon sells throughout the entire US, I should not be surprised that the data is much smoother then I anticipated – 11 PM on the East Coast is still only 8 PM in California. Without further analysis (day of the week? cd versus book? actual mailing address (therefore time zone) of the purchaser? amount of time Low Price holds?), and assuming that my Low Price will hold for a few hours, I am comfortable setting prices at 4:30 in the afternoon.
No comments:
Post a Comment