Retail data analytics: methods and applications

¶ … data collection mentioned earlier. This report shall focus on the data collection in particular. Within this report, there shall be a recitation of the data collection approach, a definition of the proper data collection channels, a description and depiction of the purchase of external data sets to assist in the problem, the development and description of data quality and other data governance issues and a planning of the ongoing storage and maintenance of the data. While planning and execute data analytics is not rocket science, as they say, it is something that is very important and needs to be done right the first time.

The author of this report knows that a ten million dollar investment is not insignificant. As such, there should be a proceeding with caution yet expediency. As noted in the prior project, JC Dollar is in the right general area of analysis but is obviously not properly discerning the effects and outcomes as is needed to craft a good loyalty. The data collection for the customer loyalty project needs to be thorough but it cannot be obvious that the data collection is going on or what it is for. Given that, the best way to collect and harvest the information is to get the information from the shopping and buying habits of the customer. As much as can be collected with no direct feedback from the customer should be the norm. Additionally, the information then needs to be assembled and cultivated in the proper way so that is fully usable and something that can be reacted to and planned for as the business and retail cycles repeat.

The involvement of third parties as it relates to data harvesting and analysis but does complicate things but it will help in terms of having a third party that has "been around the block" and has done this before. The purchase of external data sources is absolutely not a bad idea as it can reveal things that JC Dollar's data would never reveal on its own. While JC Dollar's data would be informative and helpful, getting a wider industry picture would be better. This is because some of those other stores are doing much more correctly and thus their data would be more illuminating and helpful. As noted in the prior assignment, JC Dollar's assumptions and analysis of its customer and what that should mean in terms of how the loyalty program is structured has been wrong. As such, JC Dollar should look at data sets where the reactions were correct and the outcomes were positive (Corrigan, Craciun, Powell, 2014).

However, governance of the data is important as well. For example, while having the demographics of each person in the data set is important to have, other data that is more sensitive and unnecessary to the data analysis process such as credit card information (other than whether they paid with a credit card or using some other method), name, address and so forth should be left off the data set. This is especially true of data that will be going to any third party vendors as they have no "need to know" that information and it is obviously not needed for the data analysis process and project (Arrington, 2013).

Data that is collected and saved will be warehoused in cost-efficient data storage areas. This data will show year-over-year trends and should be retained for information seeking later. That being said, one should not reinvent the wheel and start the analysis from scratch. Also, the way in which the data is summarized and assessed should be based on industry trends of successful firms and patterns within the same sector as JC Dollar. JC Dollar needs to put its own "twist" on the methods as they do not want to be viewed as copycats. However, JC Dollar should also try to not be too "cute" about the methods they employ as they do need to do what works rather than plod on with a strategy that is a failure or even unproven (Lev-Ram, 2014).

The storage of the data needs to be done carefully, however. JC Dollar should keep what they can on-site or at least within JC Dollar systems. However, the use of the prior-mentioned third party data warehousing company would be wise as well as it would be more cost-effective. Regardless, the data needs to be kept very secure and hard for non-authorized people to look at. This would obviously include people external and/or competing to/with JC Dollar but should also include people within JC Dollar that do not have the aforementioned "need to know" trait that justifies them having the access.