One of the switch engineers I worked with in the wireless telecommunications industry came out of the old AT&T landline company. While my entire career in wireless telecommunications was with digital switching, he originally worked with the electromechanical switches. The phone connection between Person A's voice would be physically connected by a crossbar to the ear of Person B.
At one time humans had to go around and collect the counts of the active circuits, a peg counter would increment and then be reset every 100 seconds. These numbers would be returned to a room full of secretaries. My switch engineer friend had a group of ten secretaries who reported to him. These secretaries would receive the counts and hand draw the charts which he then used to create his equipment utilization and traffic reports.
Slide rules, logarithm tables, books of mathematical formulas for advanced functions - this is how engineering was historically done. Yet, as computers improved, those heavy duty calculations could then be programmed and included in software packages. This increased the number of switches an engineer could manage, which increased the speed with which switches could be added into a network, which meant that customers of AT&T could have more than one phone in their house or did not have to wait 6-months to a year for their own phone number or party line.
Yet, the technology to acquire the data, transport the data, clean the data, and then visualize it was so expensive that only companies the size of AT&T could afford - not just the technology - but also the technical people to manage the data, create the analyses, and develop the visualizations. Telecommunications data, that is, information about when a handset was picked up, the numbers dialed, the numbers transmitted to the switch...all this data was one of the first industries in the world which had to grapple with the growing volume, and variety of data. Without being able to tell how much traffic was flowing through the network at any given time, telecommunications would have remained at the human operator level.
And then came the internet... and with the internet came the ability for local businesses to reach audiences outside of their hometowns or cities. Isolated government entities are now expected to make the public data public. Computing power has increased to the point where very sophisticated algorithms with millions of lines of code can run on personal machines - which means to my story - that no longer do you have to look up the number of lines a switch needs to bear the expected traffic, it can be calculated immediately. The significance of this is that those tables often limited at 100 lines, or 10,000 lines. The ability to compute the result with an algorithm meant that 1,000,000 (One Million) lines could be calculated. No one had to hand enter / hand build the tables any longer.
And now...now these capabilities are accessible to any business. Not just companies like AT&T or Ericsson. They're accessible to The Dancing Dog, a craft beer bar on the backend of nowhere. The problem is these companies don't know what they don't know. There's a revolution going on because those companies which do know that they have access to this data, that the data can have additional data added, that the data itself can be transformed... well, those are the companies taking advantage of this new knowledge. They're finding lost customers, their plotting the profitability of geographic locations, they're expanding markets. They're giving their customers products their customers adore.