Mining through systems: Facebook and Tesco

Facebook and Tesco supermarket (private companies) seem to be going in opposite directions right now. The former is increasing its share value, the latter has diminished it. Both companies work with information systems that are generating benefits but also headaches.

I am not an expert in valuing companies neither am I knowledgeable of initial public offerings (IPO). But I can venture to guess why is it that the information systems of these companies seem to be playing a key part in the companies current valuation and with it in their successes or shortcomings.

Facebook could be offering any potential clients a database of active customers who increasingly spend more time in the Facebook site. Advertising from this database could increase sales for those announcing in Facebook. With new applications coming to feed from and into such database, we now see that our preferences and those of our 'friends' could be used for those wanting to sell us things. Moreover, those individuals or organisations who want to remain connected to customers or to other people can do it through Facebook. The value of 'attention', the value of 'retention', and the value of 'networks' are now playing in Facebook's favor.

It still remains to be accurately seen though, how much conversion to sales can Facebook offer from its advertising.

Tesco supermarket is one of the textbook cases that we normally use to teach our students about management information systems and in particular data mining. Tesco was/still able to gather data from its customers, mainly purchasing habits. This data feeds into the systems that forecast demand and connect to Tesco's suppliers. A sophisticated system that enables managers to track the behavior of customers and products, and act appropriately.

As with Facebook, it remains now to be seen how their data mining systems can help Tesco 'recover' or 'bounce back' from their disappointing end of the year sale season.

One thing to consider though is that (un) fortunately Tesco and Facebook databases do not store everything about us. They do not store our worries, the tricks that we do on a daily basis to get by money wise, or if we decide to go with the competition. They can store what we fail to do (yes, we did not turn up to buy things, or we did not comment on the status of our friends, we did not upload pictures from our last trip, sorry). But they do not ask us questions unless being told to do so by a software application and the person in charge of it.

You might say that there is a problem of quality here, that without good quality data or good quality products and services to provide there is not much that data mining systems can do. But rather than focusing on 'data', maybe these systems should start asking us questions. Not the questions of the sort "What do I need to do for you to buy from me?", but "How can I help you to live your life better?", "What is going on in your life?" or by really listening to what we have to say. Here data mining can still play a role if we direct it to answer these and similar questions.