The Evolution of Data Solutions

23 May

To borrow from IBM,

“Every day, we create 2.5 quintillion bytes of data — so much that 90% of the data in the world today has been created in the last two years alone. This data comes from everywhere: sensors used to gather climate information, posts to social media sites, digital pictures and videos, purchase transaction records, and cell phone GPS signals to name a few.” In fact, we’re producing so much data that experts estimate 2016 will be the year when humans produce more data than we’re capable of storing.

With all this data being produced there’s an amazing opportunity to capture and analyze it for massively useful solutions. Solutions like tracking and curing diseases; delivering resources, products and solutions faster – for less; and perhaps most importantly, moving humanity through it’s next paradigm shift: giving us all more free time as data makes decisions for us.

But as much fun as it is to postulate when (not if) decision engines will give us more free time, we should examine the reality of data solutions for much of our business society.

The first evolution of data solutions was historical analysis. It was the capture, organization and presentation of, “the rear view mirror.” These reports would tell you what happened in the past, with little explanative power.  The user would need to interpret what might have caused the outcome. Firms would hire expensive data analysts to pore over data with tools like SAS. Forget about your average business owner having the time or money to analyze the data appropriately.

The next phase of evolution was predictive. This was discovering what happened in the past, finding trends, and stating, “if the conditions are like so, this is what will happen in the future.” During this phase the term Big Data crept into our vernacular; predictive data systems needed to pull in loads of information to improve the accuracy of its outputs - perhaps weather, GDP or other externalities that should be baked into the projections. This entailed taking lots of data from lots of different sources and letting machines help with the grunt work of prognostication.

Companies employing predictive analytics would still use data analysts to take a forecast and translate it into concrete action. It was only until recently that predictive solutions started to make their way into the middle market. The SMB (small, medium business) segment of the economy has yet to see much in the way of predictive solutions.

The most recent phase of data evolution is prescriptive analytics. Prescriptive analytics take the predictive output and recommends a decision. The decision itself is quantified and annotated so a human operator need not rummage through the data to determine what the decision is worth. In it’s most simple understanding, prescriptive analytics does all the heavy lifting of historical analysis and predictive analysis and then gives the user a digestible recommendation. “This is what you need to know from your data,” if you will.

We’ve seen firms move even one step further and let the machine make the decision for them. In other words, if the recommended action is A, and the confidence in the outcome is high enough, the machine will execute A automatically. You’re probably already using such technology but don’t even realize it! Ever booked a hotel or airline reservation? Through something called a Global Distribution System, sophisticated hotels and airlines are letting machines ingest price and booking data from around the world to determine what your booking should cost.

The majority of SMBs find themselves on the fence between nothing, and historical analytics. Depending on the vertical in question, even large enterprises find themselves with historical solutions but very weak predictive powers. Many of the legacy point of sale (POS) companies have developed fairly crude reporting to show the operator what happened last week, month and year. Cloud POS companies take the same crude reporting and make it prettier. A few Cloud companies have caught on that there is value in predictive solutions, but it ultimately gets to be a big distraction from their core business of sales, support and feature additions – despite what their investors think.

Since data science and machine learning prices are dropping, we’ve developed prescriptive solutions for the market. We see more and more SMBs using these tools every day, effectively leapfrogging the historical and predictive solution rungs of the evolutionary ladder, despite what antiquated systems their large, public competitors may use. In a way it’s not unlike the third world, who lacks basic infrastructure and are thus progressing straight to mobile for services like internet, communications, and payments.

Paradigm shifts are great for society: they advance our progress and give us more free time to discover new ways to improve our quality of life. We went from whaling for energy to drilling for it; manual farming to mechanized replacements; and from pen and paper to computers. Now we’ll go from data analysis to quantified decision making… or even automated decisions! How cool to see this next shift unfolding, and how humbling to be a part of it.