Will the telecom world focus too much on the Why, thus limiting the ability to take full advantage of the What?

Sitting in the Dublin Airport with a Guinness, on my way back from Mobile World Congress in Barcelona was as good a time as any to reflect on the show in the context of one of its biggest themes: Big Data.

One thing was certain as I walked the halls at Fira de Barcelona speaking to both C-Level and PM types, the Big Data wave is coming and it will be a significant part of telecom in the near and distant future.

My main question is, will the telecom industry internalize the trend and ride the big data wave or get stuck paddling out?

In the reading I have done around Big Data (which admittedly, has been very focused on the work of Viktor Mayer-Schonberger and Kenneth Cukier) there are three major differences between a small vs. big data world that I think should be considered in the context of the telecom world:

  • The shift away from statistical sampling to predictive modeling… In a small data world, we needed to rely on sampling due to a limited availability of information to process. In a big data world we have access to 10,000x if not 100,000,000,000x the data points alongside low-cost processors that can handle and churn them. Quickly. 
  • The shift away from focusing on the exact to focusing on the statistically likely… In an analog world where sampling is relied upon to understand the unknown, the things we can count and be exact about have a lot of importance. In a big data world, we will become more and more comfortable with messy data provided it can be correlated to produce actionable insights.   
  • The shift away from focusing on causality… Building on the two bullets above, a big data world will focus on the power of data in terms of what it can tell us vs. segmentation and why it looks a certain way. The example often given for this is the correlation between cancer patients regularly taking aspirin/OJ and the associated remission…do we really care why that is the case?

At the show, I heard over and over again: “Operators want to know why?” … “Operators have no idea, we show them” … “We help operators understand what’s going on in their networks”

Is that enough? Are we taking full advantage of the massive amount of data traversing telecom networks that we can now understand and use with today’s processing technology? Take a standard telecom example – SLA management. It’s great that today’s OSS software allows a NOC engineer at an operator to visualize and understand what’s causing degradation in the network and thus manage SLA both on the customer and vendor sides. That said, I think a CIO really only cares about that process being managed and properly correlated with BSS, charging, and other internal IT systems. Isn’t there some predictive modeling or analytics output that could be leveraged to automate and optimize the process going forward? Couldn’t the SLM actions that need to be taken to deal with an SLA issue be logged and then correlated to provide more insights around deploying marketing dollars?

I guess my questions boil down to my hope that operator mindshare (and technology budget) shifts, in the coming years, from network data collection, visualization and manual control to network data processing, self-learning automation, and service orchestration. As cellular networks and the public internet begin to look more and more similar in the eyes of mobile-focused end users, I think this will be critically important to operator’s long term survival.

To the industry vets scoffing at the thoughts above, I know I am a twenty-something that hasn’t seen the last 30 years of telecom. I also know I don’t have a deep engineering background. I also know that there is an immense amount of complexity around the protocols and required system openness that would make what I am talking about truly possible. My message and question for you smart people is: are we shooting ourselves in the foot by over rotating on the Why and not on What technologies we can embrace to make the industry prepared for the drastic shifts to come (when the Facebook generation grows up and expects cellular 2G/3G/4G/5G protocols to deliver in the same way as web service protocols)? To those still scoffing, you can reach me via WhatsApp.