The idea of “Big Data” has generated a lot of hype, especially in recent years as the variety of data sources increases and affordable new data tools appear. While success with big data analytics is not guaranteed, the chance for discovering operational improvements from your own data still makes it attractive. It’s been estimated that a large retailer can increase margins by over 60 percent through process insights. Increasingly in the social media era, big data is shaping consumer-brand relationships.
For many companies, however, big data is not providing the anticipated returns. The value in data-driven decisions seems to be elusive. Colin Strong, the author of “Humanizing Big Data“, advises looking to the human rather than the technical side to obtain positive results.
Companies need to re-examine big data operations in terms of the humans involved. This can mean not only the consumers feeding a flood of data points, but the data analysts themselves.
The Humans Behind the Data
Regarding data analysis as a statistical exercise is missing out on the human circumstances that generate the information. Human social environments can shape consumer preferences rapidly and profoundly. Basing decisions solely on the numbers means that potential changes are missed, and brands could miss spotting the next big trend or idea.
Is your business so focused on crunching numbers that real human motives are irrelevant?
Even if accurate, real-time data is limited without understanding its source. The more that human nature is excluded from forecasts and problem-solving, the more likely it is that the conclusions will be useless. A real marketplace takes direction from political and social influences, how people communicate, and how fast these relationships form and progress in a digital environment. Brands primarily need to learn how to engage audiences, further discussion, and react to both positive and negative outcomes.
Data processing can never be completely automated. A human element critical to understanding the data also needs to be considered in analysis. Viewing human activity in terms of mere numbers is ignoring the qualitative information. Decision-making as a sterile mathematical exercise is dangerous.
Every data point represents a human action and therefore involves need, motive, influences, and empathy. Particularly with modern machine learning algorithms playing a bigger role, asking WHY something happens is forgotten. If you want to know how humans behave, you have to put the measurable results in a human context.
Finding the Right Data Team
What kind of framework should organizations look for that can help them better understand the human side of analysis?
The dominance of data scientists in market analysis is part of the problem. Other important specialties like sociology and psychology seem to draw less confidence than a column of sales figures. This one-sided approach is one of the reasons that big data no longer seems to be living up to its potential.
Successful interpretations of the right channels, ads, or messages should not be the exclusive province of IT departments. Weighing human activities requires experienced observers trained in cognitive and social behavior. This can add a whole new perspective to what numbers matter, what they mean, and how they should be used. Smaller businesses may rely on their business intelligence tools for answers with no data expertise at all, resulting in a scatter-gun approach that misses far more often that it succeeds. That scenario provides little real value.
Analysts should possess the required technical and mathematic skills to refine data sets and implement analytic tools and processes. But someone with significant knowledge of the human influence must also be part of the team. Those who understand what is taking place when the data is generated are best suited to defining criteria and eliminating irrelevant facts or statistics. While some analysts are hoping for AI advancements to solve many of their problems, it won’t solve any shortcomings today, and will likely fail to provide everything successful marketing requires.
Professionals who can expose those human motives have a greater chance of discovering real insights and providing real value. They can provide the real-world context of human purchase decisions, not the endless numbers that analysts tend to sift for simple patterns that may not hold true. Genuine experts on human behavior can validate or question statistical conclusions based on real experience with market segments or pricing. An emphasis on mathematics rather than primary research such as surveys and customer feedback could be perpetuating mistakes while consumers are asking for something different.
Companies can’t afford to rely on numerical results to consistently deliver valuable insights, especially in a world where customization and innovation are becoming more essential to brand position. Those who work with data tend to trust that, given a large enough sample size, the conclusions are trustworthy. However, big data can also mean more false, outdated, or irrelevant data. More data from more sources, funneled through increasingly complex processes, also means more variation and more chance of error.
In making an analysis and providing solutions, data scientists need to think beyond the comforts of seeing data insights as completely objective and irrefutable. To understand what all this data means, analysts have to evaluate the unpredictable human natures that generate the numbers in the first place. This can be a complex process that changes from day to day or place to place. There’s always the chance that data can be misinterpreted if analysts insist otherwise. Even the most complex algorithms aren’t complex enough. Understanding human nature is never an easy task.
In conclusion, it’s important to understand that summary calculations don’t necessarily, if ever, give a real picture of human behavior. It’s important that human perspectives, motives, and context be understood before the numbers have real-world value. Even with experienced data scientists, the conclusions to be drawn are only as good as the facts filtered into your business intelligence systems. Organizations must learn to see the human influence when using big data in order for it to provide long-term value.