Let me start by talking about small carriers, the folks who read this blog. For many years we have had a good grasp of what we need to know about our customers. Companies have had somewhat flat database files that captured things like the customer name and address, a list of the services they purchase and a list of the equipment associated with their account. Some small carriers have taken this further and captured things like a history of every service call at a given address. Overall, if you consider the amount of data that we keep on any one customer it’s a manageable pile in information.
And most of the world was just like the little carriers. If you went back ten or fifteen years it’s likely that the customer database at a large carrier like AT&T Wireless was not very much more complicated than the databases at the small carriers. But over the last decade, and particularly over the last few years a lot more data about customers is available to somebody like AT&T. It’s natural that they would start capturing new kinds of data, even if they aren’t entirely sure how to use it yet.
So what kind of data is available to AT&T that was not available ten years ago? Following are a few examples:
Geospatial Data. Once GPS apps were activated in smartphones we gained the ability to know at all times where the phone (and presumably the caller) is at. In the past AT&T could identify the cell site of a caller, but only during the time that they were using the phone. They didn’t have any way to otherwise routinely know the location of a customer.
But today through the GPS built into a smartphone they can know where a phone is at all of the time it’s powered. So naturally companies like AT&T gather this kind of data, as do many other companies such as Google or Apple who control the operating systems of the smartphone. This data is just now starting to be of use in targeted advertising and is expected to become more and more valuable over time.
Voice Recognition. In the old days somebody like AT&T would have noted that a given customer called them. They would know what the call was about to the extent that somebody at AT&T took notes that became part of the customer service record. But today natural language processing technology means that customer calls can be recorded and then saved as text. AT&T can have a stored transcript of every call they receive.
Text Analysis. Whether the source is transcribed voice call, a text message, a tweet or an email, there is now software that can make contextual sense out of correspondence with your customers. This means that you can not only save text from customers but that your software can take a stab data analyzing it and categorizing according to key words and phrases within the message.
Sentiment Analysis. To go one step further, there is now software that can make a judgment about the mood and intent of customers. This can be done with recorded voice looking for emotion or can be done with text looking positive or negative sentiments.
And these are just examples of the new source of customer data that any company might have access to. AT&T is now storing a vastly greater amount of data about each customer than they did in the past. They are keeping a record of where the customer is at all times and records on every interaction with the customer.
This doesn’t even consider the huge amount of data that AT&T Wireless can derive from being the ISP. As an ISP AT&T knows every website a customer visits, the text of every email, tweet or Facebook posting, ever topic they have looked at with search engines, every video watched over the web. The amount of data gathered by the ISP side of a large company is truly daunting.
We are at the very beginning of the big data analytics industry. It’s not hard to believe that within another decade that companies will be able to accurately profile their customers (or can buy the data from somebody else to do so). It’s already eerie how much somebody like Google can know about you. The next evolution of the industry is probably going to come from incorporating artificial intelligence from something like IBM’s Watson computer platform that will direct the analysis automatically.
We all know that what the big companies like AT&T do today that smaller carriers will be doing in ten years. For example, the software to analyze things like the sentiments of customers will get cheap enough or common enough that it will come built into customer service software. And this means that your company will be gathering vastly more data than you do today.