From the early days of shorthand writing to the typewriter to the keypunch, we have been collecting data.
The accomplishments of that data give us understanding. Now we have the ability to capture data to help us establish baseline references for information processing.
Many of the processes we use to drive day-to-day activities in our society.
Government and business services, product development, goods sold and marketing are all based on data and the ability to measure data. This is important because, frankly, you cannot manage what you cannot measure. That drives the impetus towards metrics and collecting information or data that will ultimately transform into knowledge.
Once we see something happening in a cyclical way, or if we are able to find a pattern and see that pattern as repeatable with understood inputs and outputs were, we should begin to understand what was happening inside the lab box.
Within various day-to-day operations as organizations, we have been able to infuse and conjoin manual processes with digital, automated processes. Such is the same for manual data entry and automated data entry. This transfer into a more automated, data-driven pathway is starting to open up insight at a faster pace because the data itself is allowing us to study and analyze it. If you ever talk to anyone developing a machine learning platform, they will tell you they need lots of historical data.
Many of the insights from the future are based on studying the past.
When we go through history, we are able to say, for example, that on May 5, 1995, we captured a lot of data on that day from newspapers, videos, audio recordings, letters, and emails, etc. That becomes a rich platform because we also know what happened on May 6. And because we know what happened on May 6, we are able to begin training a computer through a model of algorithms to match the outcomes that we captured on May 6.
In other words, if I am able to take the data from May 5 and all of the info in a 24-hour period and put that into a mathematical algorithm with different models and identify different data attributes, then I will be able, over time, to gather the right representation of the data and event so that what comes out of my model is exactly what occurred on May 6.
Now I know I have a machine-running platform that is beginning to tune into and train to start providing the output that occurred. Mind you, this is using past data, not any future information yet. This is the beginning stages of neural networks and artificial intelligence and analytics.
Now we’re moving into a world where this is happening in near-real time.
Where we are able to take the algorithms that have been tuned and trained over the years and build a level of pattern recognition and predictive outcomes that utilize the past history as a baseline, and then look in the moment at what comes out of the predictive algorithm and what is happening in near-real time.
That allows analysts to shift and create insights from a year’s worth of data to near in-the-moment data applications. This is where something like fraudulent protection occurs.
Take, for example, an American Express card swiped at the point of sale to complete a transaction. American Express has the capability to run analytics to get insight into the nature of that transaction: is it fraudulent? Is it seemingly fraudulent? Is it a valid transaction?
The insight they gain from that particular transaction allows them to manage the risk against that transaction. It either becomes a profit or a loss for them based on their ability to tune their system in a way that’s almost in real time.
In order to protect themselves from fraudulent activity, organizations like credit card providers had to come up with a way to take insight in near-real time to stop the risk of losing money on a transaction.
To do that, they have to leverage data, and the data becomes very rich. Within 100 ms, they have to make a decision on the validity of the transaction. They get there by leveraging data.
The data that they’re capturing allows them to take information and put together relationships between the various data collected. The first step is to get access to the raw data so that it can be monitored. I believe that you cannot manage what you cannot measure, but you cannot measure what you cannot monitor.
Monitoring is where the data input and data capture occurs. It requires exposure to data we produce every day. Mobile devices are the most valuable devices to have as a human today. They provide one of the richest data sets for every organization in the digital age.
We do so much activity and perform so many functions on a mobile phone. Because they are physically tethered to our bodies in some way at all times, they become an extension of us, giving insight to our location, relationships that connect whatever happens on apps, conversations on the phone, etc. These all create relationship insights between the data points coming into a system.
That insight allows data miners and processors to have a more accurate depiction of what the truth is and trigger a response to that action. If the truth is that I am actually using a fraudulent credit card, the insight gives the credit issuer the opportunity to shut that down; to do so at the transaction level means the goods and services rendered won’t be disrupted by a fraudulent transaction.
When you look at where we’re going in this digital age, you have to look at the risk associated with every action and transaction.
You’ll begin to understand who’s motivated to collect data and drive insight from that data in order to address that risk. The least amount of risk you have in day-to-day transactions and activities, the less motivated you may be, outside of pure curiosity, to move toward a digital world.
The more risk I hold and am responsible for, the more incentive I will have to drive towards a digital outcome or adoption because I will have access to data that would normally never be available to me as an organization, entity or person. Once the data is collected, it has to become actionable by performing certain functions on the data, whether that’s modeling, analyzing or predicting the next outcome based on the current and historical state.
With all of the benefits that insights bring, it also brings disruption to other processes, technologies and ways of living and working that is no longer going to be relevant in a digitally driven society. If I’m going to stay relevant going into the future, I should begin to understand how to collect insights on myself.
Google provides personal analytics. You can go to a certain section of Google and it will tell you all sorts of things that you’ve done: what sites you’ve visited online, where you’ve gone, and that’s your personal analytics dashboard. In order to give you insight, yes, they’re collecting that.
Whatever the trade-off is in terms of privacy and approved actions based on insights Google is taking, that’s where our society has to begin formulating new laws and policies on privacy.
At the end of the day, though, that insight is very important. It gets more interesting when we start developing ourselves for the next layer of digital transformation and understanding the cognitive patterns that we generate.
These patterns will answer questions regarding what we think about and discuss most, as well as our greatest fears and aspirations.
Once we start to leverage that data, we can start to turn that towards ourselves.
This is cognitive thinking as a service, or cognition of AI that is getting collected that allows us to get insight into ourselves. What are my strengths and weaknesses? What are my professional capabilities and potential based on what I’ve done in the past? All of that becomes available to us as well.
This may not sit well with average citizens. They may not want to know everything about themselves, or they may not be comfortable with some of the insights gathered. That could be disruptive to their own psyches.
This is the paradox hovering over society when it comes to insights. We enjoy the fruits of insight from a protection standpoint (cybersecurity, national and personal security); we appreciate leveraging multiple data sources to get insight into our commutes so that we can avoid delays.
But the more we consume, the more we actually feed the same system we use, ultimately resulting in our feeding the data continuum.
We’re right there in the crux of trying to figure out how to take full advantage of technology but also have the ability to go off the grid when we want to. This can be difficult to process, especially when considering the fact that most are not privy to anything that data collectors are doing.
Understanding the digital age and what it means is really important for every professional at every level of every occupation. The discussion around that has to occur because there has to be an understanding of what is palatable and what is not. This is the question that is beginning to resonate in the marketing sector because that’s where much, not all, of data collection, is occurring.
Market data is where we’re being exposed as individuals. If you look at the ad technology market or the ad tech industry, you’ll see that there are many insights into the predictive nature of a person based on similar analysis of what is being posted, particular locations and the time that data is being shared. All of that develops a digital persona.
We now are in a place where we have to make decisions as members of society regarding how far we should take this, as well as what to do with various insights once we get to a place where they can actually predict what we’re going to do tomorrow based on predictions of future events. Are we there yet? Of course not. Are we on a path towards that type of innovation? Definitely. The insight continuum, from what I’ve observed, is just beginning, and I don’t predict any foreseeable end to it in the future.