Exposing Under-Performing Products

With 1500+ products it is very easy to list an item and forget about it, especially if the item never sells. This program is designed to bring these products to my attention.

This program looks at items listed in the last 30 days and tells me the following: total purchases, days listed, and percent of total quantity sold. I sorted so the oldest items with no sales come up first.

What can I do with this information? The best thing I can do is adjust their titles using a rough approximation of eBay's Best Match algorithm. (A guide to this poor man's SEO on ebay can be found here).

When that doesn't work, I cut the price. Identifying and liquidating under-performing product is just as important as making money off the better items.

Program logic

In order to analyze clothing sales I have to compensate for varying item titles. A normal item title looks like this:

ADIDAS CLIMALITE COACHES POLO NEW WHITE GOLF XL

I used the substr() function to select only the first 40 characters of my title. While it isn't pretty, I get the information I need to identify the slackers (and the go-getters, which may or may not be underpriced). Nesting TRANWRD and SCAN functions identified and removed size, but also removed the size letter elsewhere (So a large (L) OAKLAND RAIDERS item came out as OAK AND RAIDERS, and padding with a blank only works when the size letter doesn't start a word).

Once the different sizes of an item have the same title, I use first/last processing to sum their sales and quantities.