Time for Fall Cleaning
Growing up in an Italian family, the ritual of spring and fall cleaning seemed to respectfully greet the approaching season. Spring hinted at the upcoming summer, a time to put on your shorts, picnic, and re-engage with friends. Fall on the other hand, signaled cooler, shorter days were ahead. Wool sweaters came out of storage, heavier drapes were re-hung, and carpets were put back on the floor to warm chilly feet. This daunting task of cleaning was lessoned since all of my Aunts pitched in to help each other get their houses cleaned. Judging by their laughter, they seemed to enjoy their time together. And who doesn’t enjoy a clean house!
Fall is also a good time to clean your mailing lists as many venues gear up for a busy season.
Typically, data gets dirty because transactional purchasing databases are designed to connect online sales via a shopping cart, retail sales Point of Sale, to credit card processing and inventory. However, these portals often lack the additional software that automatically corrects addresses, including email addresses, during the input process. Without address correction software, incorrect data eventually contributes to data quality problems.
Some of the most common input problems we encounter are inconsistent formatting, foreign addresses, carriage returns, Cyrillic scripts, false addresses, house accounts, unknown codes, and missing data. Our typical processes are:
- to run scripts to remove the show stoppers mentioned above;
- household the data (meaning putting together all the data for a household);
- standardize the address, city, state, and zip;
- append if necessary to increase insights;
- visualize the data into categories such as income, gender, education etc;
- model the data.
Cleaning the data takes the most time in the analysis process. Based on the amount of data, it can take days to months. If you don’t take the time, you may have incorrectly analyzed a problem or solution. This will surface in all sorts of ways from the wrong product fit, incorrect pricing, low marketing campaign ROI, and brand defection. Simple, Garbage in will produce garbage out. Good examples of product failures are Google Glass, HP’s Touch Pad Tablets, and Microsoft Zune.
If you do not have a data strategy in place, now is a good time to create one with your cross functional teams. Be sure to include:
- Data input guidelines
- Data refresh schedule
- How to handle incorrect or dirty data
Similar to my families cleaning rituals let PatronLink do your heavy cleaning…of your data. In fact, it will cost much and be cleaner than hiring any house cleaning service. While we do the work, you can enjoy a festive pumpkin drink at your local coffee shop.