Throughout the history of digital technology, data was largely the language of vigorously trained computer scientists and engineers. Only in the past decade, with the influx of highly-targeted social advertising, has data really entered the public consciousness. Privacy legislation like the European Union’s General Data Protection Regulation (GDPR), enacted in 2018, and next year’s California Consumer Privacy Act (CCPA), have helped data become part of the common tongue.
This legislation means more control for consumers and accountability for enterprises, ultimately laying the groundwork for data’s next phase: democratization.
Utopia In The Enterprise?
When I first heard this idea of data democratization a few years ago, it was said that the democratization of data would create a technological utopia. Of course, utopia is elusive, and societal applications of technologies — however well-intentioned — do not always pan out as planned. At first, many believed Google’s search engine would usher in a sort of knowledge utopia, but we’ve seen how this ideal can be corrupted when applied on a global scale.
However, unlike a search engine, which is used at a truly unfathomable rate for every purpose imaginable, the democratization of data will have a more focused impact on the way businesses and their employees access data and use that data to make enterprise decisions. Currently, data discovery practices are largely cumbersome and antiquated. It might take months to go through proper protocols and receive requested data from the IT department. This, frankly, is too slow. One of the ideals promised by the democratization of data is that all business intelligence would be accessible to each employee in just minutes.
Following The Path Of Autonomous Cars
This all sounds nice, but how close are we to the promised land? It’s helpful to compare our progression to that of self-driving cars in recent years. At first, we had driver assistance: the car would alert the driver if it thought they might be dozing off at the wheel or drifting out of the lane. Then came semi-autonomous driving, which allowed us to let go of the steering wheel while the car drove itself for periods of time. This is akin to today’s supervised machine learning technologies. The final stage is fully autonomous driving, in which we can simply ask the car to take us where we want to go.
Most of the widely used enterprise technologies today fall into the first two categories. We have engineered various business systems and databases to communicate with one another so that data can be more easily filed when it arrives in the new system (assistance). We have also begun automating the actual transfer of the data between workflow systems in a semi-autonomous way.
If you’re following this analogy, you may be thinking that the business equivalent of the fully autonomous, driverless car is a business analytics robot that has taken the place of a human worker. In both cases, however, the human is relieved of a mundane task that had consumed much of their time and attention, freeing them up to think critically while the mundane processes continue to run in the background.
Data Democratization In Practice
In June, researchers from the Massachusetts Institute of Technology and Brown University previewed how “driverless” business analytics — officially known as automated machine learning (AML) — might look in practice. For the past several years, the researchers have been developing a drag-and-drop analytics interface called Northstar, and a recent update allows users to instantly become virtual data scientists, applying complex machine learning tasks to large data sets by tapping and manipulating modules on a touchscreen.
In a mortgage approval setting, for example, an underwriter could use this technology to not only cull the data for the particular loan applicant but also have the ability to immediately and accurately compare that data with local mortgage approval trends, as well as other approval decisions he or she had made in the past. Likewise, doctors and physicians could easily compare a particular patient’s test results to entire databases full of (anonymized) health data and tell the machine to identify notable patterns that lead to smarter treatment.
AML technologies like these are still largely in their infancy, but business leaders must understand that they are coming. IDC research published in March predicts that global spending on artificial intelligence this year will rise 44% this year alone, with that spending nearly doubling by 2022.
The adoption of new business tools must be coupled with a concerted effort to educate employees so that the technology is not a wasted investment. But, it’s especially significant in the context of the democratization of data because this technology will, in theory, be available to every employee — not just technically trained analysts. Fortunately, the aforementioned privacy legislation has introduced a base level of data literacy to the general public, flattening out the learning curve.
Don’t Lose Sight Of The End Goal
In these data training endeavors, keep in mind that while great changes like these will have the most drastic impact on employees, the tech executive’s mission must remain steadfast. In company-wide communications, we often compromise our messaging for fear of being repetitive or boring. But, part of successfully introducing new tools to the organization is hammering home a key message. What’s exciting and perhaps challenging for staff should be consistent and perhaps boring for business leaders.
This future of democratization offers a look at data’s lasting and maybe even final stage. What began as the language of an elite, select few eventually became less esoteric and more accessible thanks to technology. Today, data is transforming from a common language to our universal language.
