I attended a talk on Social Media Analytics yesterday, organized by the Institute of System Studies (ISS) of NUS. As usual, it was well organized and the hosts, IBM, made us feel very welcome. Inspite of no longer being involved directly in this space, a number of companies which my employers have an interest in, do, therefore, I still try to keep current as to what is evolving in Social Media.
What disappointed me greatly about the speakers (and this is not really their fault), is that the systems they were presenting seemed primarily to be Social Media monitoring and aggregating tools. Having run a boutique Data Mining consulting firm, I take the term Analytics to mean that insights are developed from the data without being directed by a user, or at least preliminary conclusions can be drawn which can be further tuned by the user.
The fact that there is no such tool for Social Media may be less a matter of technology than it is of need. Monitoring Facebook or Twitter isn’t going to help a company understand more about their customers who buy for example, cereal. Because the final sales transaction (even if its through an ecommerce site) is concluded outside the Social Media space, it is difficult to draw a correlation between Social Media profiles and customer profiles. Many companies have found to their detriment that lots of Social Media chatter about a new product does not necessarily translate into outstanding sales.
The one area where Social Media monitoring remains a good bellwether however, is where there is negative criticism online. The reality is that people are cynical animals and reading how good something is will still give them pause to consider if that something is suitable for them. However, if they read bad reviews of that same something, the chances are far higher that they would no longer consider it. So companies do have to monitor the Social Media space for negative criticism.
If there is a Social Media campaign behind the launch of a new product then monitoring will of course help a company understand how effective it is in terms of reach, however, unless it is possible to segregate or stagger the timing of the online launch and real world launch, a true correlation about the actual success of the Social Media campaign and actual sales, will be hard to draw.
You will of course say that there are product categories that are exceptions to this broad generalization above. Handset makers, consumer electronics, car manufacturers, may be given as examples of industries where Social Media activity and actual sales can often be better correlated, however, these industries tend to spend widely across all marketing channels not just online.
So while I still advocate the use of monitoring tools, due to the anonymous nature of Social Media and its tenuous links to the final transaction that defines the business, analytics in this space is a waste of money. I am well aware it is possible to piece together detailed information about a person’s profile online, but why go to all that bother (and possibly face charges of being intrusive), when a credit card transaction not only validates that a sale has been made, but also provides all the customer profile and information in the system?