To me analytics has been around for long
In the 70’s
Airlines were “computerised” way back; I flew in 1976 from Madras to Chicago using a “computerised” ticket!
Out of the Airline computerisation came “planned overbooking” – Airlines found a “pattern” – last minute cancellation peaking on weekends but near zero on Monday morning
Simply the weekend effect and Monday morning catch up; thanks to data stored on tapes they could estimate with near 100% confidence, possible “cancellations” and introduced “planned overbooking”
It helped Airlines and passengers – but not scaleable – at best 20 Airlines in USA and may be 200 in the globe (at least then)
in the 80’s
Any Data mining book would give the example of another “pattern” – men in 20’s & 30’s picking up Diapers AND Beer cans on Friday evenings; by putting them together they could increase sales
Something that can be scaled to thousands of stores
In the 90’s
ATM switches could see a pattern of “unlawful” ATM withdrawals; by people who “stole” ATM PINs from unsuspecting customers
This was scaleable to millions as ATM base was nearly a million
data mining software could “discover” pattern; for example,
using software “IIMB student store data” could throw out a pattern; 89% of purchases happened in 2 days!
Everyone knew that most students purchased their items (tooth paste or pencils) for the full 3 month term;
the discovery was the extent 89% it is better to ask some stores to put a “window” twice a year rather than running a store all thru the year!
That is “discovery” and the area was called Knowledge Discovery; even today KDDI is the Conference for data mining!
What is possible today is to use Analytics to “disrupt” an industry c
For example, an industry called taxi hailing is getting disrupted right before our eyes; our own TaxiForSure is doing in for Autos in Bangalore since yesterday!
Uber started it in SFO 5,6 years back
Thanks to GPS, Smart phones, Google Maps, you can bridge the “information asymmetry”
For example 90% of the time a cab / auto driver has a customer within 100 feet, but the driver does not know;
Similarly, 90% of the time there is a taxi / auto within 100 feet but the customer does not know!
If this asymmetry can be “managed”, we can reduce the number of taxi / auto on the road by 30%
We can do similar exercise for buses, ambulances, fire engines,… as well as doctors and hospitals, patients and nurses, … the list can go on
That I think is the real power of Analytics
(My Pep talk on the Inauguration of 3rd “Analytics Program” (Continuing Education Program) @ IIIT-Bangalore on January 17, 2015)