Does Bigger Data Lead to Better Decisions?

The tension between science and emotion continues to influence managerial decision-making and the outcomes that flow from it.

As a society and an economy, we purport to respect information, its distillation into knowledge and even its occasional reduction into wisdom. But more data, selectively chosen, has also served to heighten our emotions as the information served hardens previously held opinions or is rejected when deemed contrary to more deeply held beliefs.

We know that bigger is not necessarily better. But what enterprises attempting to manage the new-ish era of data abundance are beginning to learn is that the evolutionary advantage provided by diversity applies to information as well as biology.

Important industries like finance and pharmaceuticals have had access to big data for years, long before it became a thing. That has not stopped the financial services industry from creating crises or even from preventing failure. Really smart people backed by substantial resources and exceptional experience continue ignore information that contradicts their world view or interferes with attaining their goals.

The heterogeneity of sources and of those who interpret them may be as important as their accuracy. Just as the notion of five year plans, once considered essential, are now regarded as laughable, so the accretion of data for its own sake may soon be scrutinized warily if the processes by which it is vetted and applied do not produce appreciably superior outcomes.

Better decision-making, however an individual or institution may define that, is probably the result of a combination of factors that could be unique to that entity. More data may provide an advantage, but only if the processes by which it is evaluated are subject to constant improvement and the people who oversee such procedures are themselves conscious of their limitations and of the need to challenge their own assumptions.

We are, to some degree, in the panacea business. We want more power, more convenience and more success. But we are delusional if we think that any one tool or path will help us achieve those goals. JL

Theos Evgeniou, Vibha Gada and Joerg Niessing report in Harvard Business Review:

Big Data can lead to Big Mistakes. After all, the financial sector has been flooded with big data for decades.
Many scholars, from decision scientists to organizational theorists, have addressed this question from different perspectives, and the answer, as for most complex questions, is “it depends.”A large body of research shows that decision-makers selectively use data for self-enhancement or to confirm their beliefs or simply to pursue personal goals not necessarily congruent with organizational ones. Not surprisingly, any interpretation of the data becomes as much an evaluation of oneself as much as of the data.
How can organizations avoid such pitfalls and turn “Big Data” into a safe opportunity? Decade-old research provides some pointers. It is not Big that matters, it is Diversity that matters. Big is old – retailers and financial institutions have had big data for decades.
But Diversity is new. Take large retailers. Sure, they have had enormous databases for long time now. But marketers are only now connecting data from loyalty programs in physical stores with data not only about how the same customers behave on the company’s website, but also how the same or similar customers anywhere in the world behave on other websites – ranging from news sites to car sites to movies sites – all tracked using cookies. They can then link this data with in-depth market research as well as social media data from Twitter or Facebook.
This kind of linkage is reaping rich rewards.  A leading Telco company we have worked with was able to increase market share by more than 20% in some countries without increasing the marketing budget by leveraging behavioural and transactional data from social and general media.
Some innovative companies are connecting data traditionally used by banks to assess the credit score of loan applicants with information ranging from mobile phone usage data to online social media relations data, in order to better and faster assess the creditworthiness of a micro-loan applicant. What can a phone bill tell us about the chances that someone will repay a loan?  Or even about the creditworthiness of the people that the applicant is connected to online?
Of course, management scholars and practitioners have long recognized the benefits of diversity.  It’s widely accepted that heterogeneous teams are more creative than homogeneous ones. Diversity, if managed well, yields divergent thinking and the pooling of a broader base of knowledge results often in better strategic choices.
The point we stress here is that diverse data confers similar benefits. And it’s worth noting in this context that in statistics and data science as well the key quality measures in data are not the size of the dataset but metrics like variance and entropy, which effectively capture the data’s diversity.
So perhaps we shouldn’t be talking about Big Data making decisions better, but about Diverse Data connecting the dots using new technologies, processes, and skills. We need to connect the dots or we risk drowning in Big Data.