Can Businesses Use Credit Card Data For Customer Analytics?
Yes, however only limited parts of the data is allowed to be stored and therefore functioned on. This is a really great question that has some rather unexpected turns. Analyzing Data Is One Thing, Fully Acting On The Insights Of The Data Is Another In the late 1970s large retailers began to use mainframe computers to track and store transactional data. In that era there were dozens of consulting companies, computer time share services, software companies and hardware companies that produced a great deal of research about how potentially powerful this data could be. Large retailers began to upgrade their old cash register to early versions of the POS systems we see today. I have some of the original research (is not on the Internet) from this era and it is fascinating to me how detailed and advanced the thinking was. In fact some of the research is still not being applied and used even today. As mini computers replaced the mainframe systems, transactional analytics became more common. However there has not been a tremendous amount of meaningful advancement since the late 1980s. Certainly the process has become far more sophisticated as well as very detailed graphical insights. The supply chain management and inventory management systems are of course the backbone of all large retailers. However some of the promise of the early research has not materialized. Individualized Transactional Data The aspect of this data could impact realtime customer interactions were thought of in the 1970s and only just lightly touched upon by companies on the Internet like Amazon. Target in my view has the proper systems in place to fully utilize transactional data and function on it in a real time environment, but they have not done this yet. Few other large retailers seem to see any value in individualizing transactional data other then to power a simple Loyalty program. Loyalty programs as they stand today are but a one dimentional shadow of what they can be with the proper analysis and real time actionable system infrastructure. The potential of individualizing transactional data has also been easily available to small and medium sized merchants since the late 1980s. Point Of Sale systems that captured Payment Card magnetic track data actually stored all transactional records in open files on the hard drive (a bad idea for security with 20/20 hindsight). Payment Card Processors have also made this data available electronically and later online in the 1990s. Some POS systems for small merchants could uses historical data to notify the merchant of a regular customers, even though some think this is a idea born in 2012. All of this data has been available for quite sometime including individualizing transactional data in real time. What Data Can Be Stored From A Payment Card Transaction? Since the formation of the Payment Card Industry Council (PCI) to help address the growth of security breaches when Payment Card data is involved. They established and continue to establish standards on what parts of account data can be stored and transmitted.
We can see in the chart above that as it stands today, to be 100% PCI compliant a merchant can only store the Cardholder Name and in a very limited way the Primary Account Number along with the Expiration Date of the Payment Card. Of course all of the merchants internal data could also be stored from each customer's transaction. Thus any merchant could certainly have rich and detailed access to individualizing transactional data and function on the data from a "big data" perspective and from individualized transactional data perspective. There have also been quite a few attempts by some POS companies to make access to this data very easy for small and medium merchants. Data, Data Everywhere And Just A Drop In Use Thus we have a perplexing and confounding issue. For example we can see one aspect of this issue with Intuit and Quickbooks. From the outside and even the inside of Intuit there is belief that the chart and most of the reporting features are being used to a high degree by merchants. It turns out in my empirical studies (16 of them), this is far from the case, in fact, the presentation of the merchant's own sales data in dashboard form and in individual charts is one of the least used features by real life merchants in real life situations outside of the view of us Technologists and Data Scientists. All this data and most merchants are not using it, why? Well that is the billion dollar question, and I would love to share deeper insight under the correct circumstances. It is not because the images need to be rendered with better colors or more creative roll over effects. I can say that the issue is just 50% based on producing really great technology and software and frankly this has not really been done yet. However just as important is the understanding the actual merchant from a anthropological perspective with a deep understanding of each MCC (http://en.wikipedia.org/wiki/Mer...) and how these business not only obtain but sustain customers. This is not and can never be a one size fits all solution. Nor can it be addressed from a top down approach of engineers and executive assuming what merchants need and want. Data and analytics are utterly useless, unless or until one can obtain the value in the correct frame of reference. This takes fine craftsmanship of technology but it also takes empirical experience of the merchant frame of reference. Guessing does not work. There is an ocean of opportunity in individualizing transactional data however only a few drops ever get used. There is absolutely no doubt that at some point a few startup companies will get it right someday. I suggest that it would happen faster by understanding history and not go about reinventing the wheel.