When Gartner hosts its Business Intelligence and Analytics Summit in Munich next week, an international community of data professionals will gather to take stock of the innovations, challenges and risks characterizing the Big Analytics landscape today.
Next week’s two-day conference Oct. 14 and 15 will offer a mix of Gartner research, panel discussions and real-world case studies, aimed at bringing together BI and analytics thought leaders, enterprise architects and senior IT professionals to explore the growing sense that big analytics are starting to create big problems.
(Editor's Note: Digital customer analytics will be discussed at CMSWire's DX Summit 2015 conference on Nov. 3 and 4 in Chicago.)
When Bigger Isn’t Better
When big data meets big analytics, what could possibly go wrong?
By 2018, 50 percent of business ethics violations will be directly traceable to the improper use of analytics, according to Gartner Research Director Alexander Linden.
That’s because as analytics applications become ever more sophisticated, their ability to dig deeper and mine hidden information inevitably leads to dark corners that lie beyond the reach of traditional analytics ― and existing privacy laws.
Penny Wise and Pound Foolish
Linden’s goal is to sound the alarm that the ethics concerns raised by Big Analytics have already gone beyond the hypothetical. He reels off lost reputations, competitive disadvantages, wasted resources and compromises to revenue potential as several very tangible downsides to unbridled analytics use.
Oh yes, and legal sanctions.
Linden points out the irony that while organizations don’t blink at investing $1 million in a top-end SAP or IBM analytics product, they often hesitate to pay the $150,000 salary of a data professional with the expertise to navigate around ethics problems and keep them out of trouble.
Like Moths to Flame
And the ironies continue.
BeyondCore CEO Arijit Sengupta points out that despite the risks, enterprises are actively procuring the very types of analytics that can create ethical problems.
Today’s next-generation prescriptive technology “is all about … finding patterns and then trying to figure out … the best actions to take to positively influence those factors,” explained Sengupta.
Unethical or Illegal?
Sengupta and Linden agree that the risk of organizations becoming so fixated on gathering customer information that they fail to consider whether they should use it has already become very real.
Gartner’s Linden cites the example of a major bank in the Netherlands that was harvesting its customers’ checking account data and using advanced analytics to identify other products that the bank could cross-sell to them.
While not illegal, Linden strongly believes that such actions sully brands and ultimately hurt business.
One Among Many
Linden notes that today’s predicative analytics applications are so sophisticated that they can extrapolate individual customer journeys based on aggregate behavior.
“What is striking about [these data sets] is that even if I opt out of everything and avoid any contact [with customer touchpoints] other customers who have not opted out will [still reveal] everything about me,” Linden explained.
Don’t Trust Your Calculator
“Data is often about things that people don’t necessarily offer knowingly, so with every click at every customer engagement point, customers are revealing information that they are not aware of,” Linden cautioned.
The “always-connected” world also poses a major issue because even when mobile apps are not being used, they are trading information. Even your calculator is communicating behind your back according to Linden.
“People are using these low-cost apps without [realizing] that the app is using and trading information from your phone,” he warned.
Avoiding the Quicksand
On the eve of its Summit, how do Gartner’s experts think analytics practitioners should proceed?
Gartner’s Linden has plenty of up-front caveats: “We can’t come up with a precise framework or set of rules for operating in this context and [promise] that if you comply with them you will be fine, he admits.
“We cannot prescribe ethics; we cannot prescribe success and what that means,” he continued.
Common Sense Guidelines
Gartner has developed a set of common sense reminders and best practices to guide organizations away from the big analytics ethical cliff:
- Organizations should seek to anticipate problems before they arise by setting out their business goals before launching their analytics.
- Remember that big data sets do not necessarily require big analytics. Complex questions, not big data sets, require complex algorithms.
- Not all problems require analytics at all. Consider other methods of gaining insight such as statistical modeling, data mining and machine learning algorithms.
- Make sure that your organization has the resources to use the analytics capabilities it buys. Remember that sophisticated analytics require sophisticated controls.
- Analytics cannot, on their own deliver transformational results. They may be better deployed to incrementally update or extend existing business processes
Following the Leader
Gartner is a leader in the movement to rein in and channel Big Analytics but it has support from other industry players.
The Data Science Association has established a Data Science Code of Professional Conduct to help organizations and data professionals manage ethical situations.
The Digital Analytics Association has drawn up a Web Analyst's Code of Ethics.
The American Statistical Association has gone a step further by creating an enforcement mechanism to apply its ethical guidelines.
With its upcoming conference and beyond, Gartner’s Linden wants to make one thing clear: Both organizations and their customers benefit from generating additional solutions to big analytics issues and moving decisively to incorporate them into their overall business strategies.
Title image by Alejandro Escamilla.