Thursday, Oct 15 2015

CIOs Must Get Familiar with Analytics

Written by

CIOs Must Learn the New Math of Analytics

Big Data is expected to bring companies big profits, but it can also bring big headaches. The value of big data comes from analytics, algorithms that look for patterns in the data. Knowing those patterns, companies can sometimes gain competitive advantages. But is acting on those patterns legal? It depends.

One example of this kind of risk is the recent news that Google's advertising systems tends to show high-paying jobs to men more often than to women. There clearly isn't a discriminatory intent, but there is a discriminatory effect anyway. In another example, an algorithm used by Athena Capital Research led to the SEC fining the company for illegal market manipulation.

Identify Risks With Using Analytics

In order to understand the risks of their analytics, CIOs need to understand the math behind them. Engineers can build systems that crunch all kinds of data, but the CIOs need to decide which data is ethical to use, how using it ties in with the company's business strategy, and how to monitor the algorithms to make sure they're in compliance with all relevant regulations.

Identify Opportunities by Using Analytics

To find the opportunities for algorithms, businesses – and CIOs – need to start with a business problem first, not the data. Once the problem is identified, you can select the data to use to solve it. By controlling both the data and the problem the algorithm solves, issues like discrimination can be prevented.

Understand How Your Customers Will React

Another usage of algorithms is to personalize the information presented to your customers. Some customers will want this; some may find it creepy. In some cases, it can feel like a violation of privacy.

Understand Public Relations Consequences

Algorithms sometimes drive pricing decisions, with consequences that go beyond their impact on the company's bottom line. Some companies modify prices based on information such as the user's zip code. This can be considered discriminatory, like redlining, as well as becoming a public relations problem. While Uber's legal surge pricing has been widely criticized, it developed into a major public relations disaster when it automatically kicked in during a hostage crisis in Sydney, Australia. When algorithms run the business, the business – as well as the public – can end up losing.