How do you know when you need a data warehouse, a data mart or at least an extracted set of files for marketing analysis? If the thought of pulling together any or all of the following analytics projects in the course of one day makes you want to throw up your hands in exasperation, then you’re a prime candidate:
Task No. 1: Run a report of how many unique 12-month buyers you have. Try this in a leading enterprise resource planning (ERP) system, and you will arrive at a number. But do you have a way to check if each record is a unique account at a single, physical address? Are there instances of multiple accounts for the same business—e.g., one for credit sales, one for purchase orders and a half dozen because your sales force spells company names creatively? Is the only solution to dump a comma delimited file into a spreadsheet, then manually sort, compare and summarize the records? That’s a sign you need analytics help.
Task No. 2: Append trailing, 12-month sales of all products and of your top product category to your list of unique buyers to do some share of sales calculations. You likely will have to write a query to extract the right date range of order detail records and then run it twice: once for all items and again for the top category. Can you merge those extracts and perform your calculations? Or, does your method involve extracting the resulting files and merging them by using a relational database package or a lookup function in your spreadsheet? This is another sign that your needs exceed the capabilities of your operating system.
Task No. 3: Create an RFM score of your 12-month buyers. Calculate the life-to-date value of those buyers segmented by the product category of their first purchase. Now tear your hair out as you devote precious hours to numerous data processing requests.