Big Data & Analytics Heroes
Bob E. Hayes
Bob E. Hayes, chief research officer at Business Over Broadway and this week's big data and analytics hero, discusses his own big data learning curve. Looking ahead, he cites the critical need for statistical literacy to become widespread if society hopes to achieve the full value of big data and analytics.
When getting started with big data and analytics, what were your biggest challenges?
The biggest challenge for me was simply to get an understanding of the area of big data. When I first heard the term in late 2011, no standard definition of it was in existence. For example, check out these more than 40 different definitions of big data. Apparently, big data can be boiled down to these six areas:
- Processing the three Vs
- Analytics approach taken
- Determining the reliability, validity, usefulness and value of insights
- Integrating disparate data sources
- Communicating the results
- Consideration of security, privacy and ethics concerns
Perhaps because of my educational background in statistics and research methods, I think of big data as being less about the data and more about what you do with the data. Therefore, my interests in the field of big data tend to focus on how data is used to draw insights. I write primarily on the topics of analytics, data integration, data science, measurement and reporting.
How did you get organizational support for your big data initiatives?
Researchers have shown that success of big data initiatives is not only about the technology, but also about organizational qualities. Based on these research findings, I encourage my clients to keep the following in mind to improve their chances of garnering organizational support:
- Find an executive with an analytics mindset, and gain the executive’s official support toward a big data initiative.
- Align your big data project with a current company initiative, mission or value. Executives cannot ignore something that supports the company mission.
- Identify a business problem and design a study to answer it—for example, will a new marketing approach increase sales? Keep the study design as simple as possible.
- Have a specific question you are trying to answer. If you can answer your question using a two-way comparison, don’t try to design a more complex study.
- Be transparent and clear about your data and how you arrived at your conclusions.
Taking a long-term perspective on big data initiatives, I encourage my clients to start educating all employees on basic statistics. Statistical skills help people learn from data. I see vendors touting their visualization solutions as a way to democratize data science throughout a company. Giving people data visualization tools and calling them data scientists is like giving people stethoscopes and calling them doctors. Having knowledge in statistics can improve how the data visualization tool is used as well as how the end consumer interprets data-heavy reports.
How have big data and analytics impacted the way you do your job today?
In big data’s wake, two areas of technology and application have directly impacted my work. First, advancements in machine learning have improved how efficiently I am able to get value from data by shortening time-consuming data management steps—for example, capturing, cleaning and integrating data. Second, natural-language processing (NLP) has expanded my thinking on how to measure attitudes. This new knowledge inspired me to improve how we measure customer attitudes and sentiments.
Do you think big data and analytics will handle the data growth in 10–15 years, or will we need another shift in technology? Why?
I’m not a technologist, so I’ll leave the prognostication to them. But based on Moore’s law of exponential growth, I think that we will likely find a way to handle the growth of data. Recall that, in 30 years, we went from walking punch cards to and from the computer lab to instantly accessing the world’s information from the palm of our hands.
As I alluded to earlier, what I think will limit the value of big data and analytics is the statistical literacy of society. Statistics is the science of learning from data. Statistics and statistical thinking help people understand the importance of data collection, analysis, interpretation and reporting of results. Statistics helps people make sense of data. It is the language of data. Because we will live in an increasingly quantified world, people will be further inundated with data-related metrics and information.
Think of the data being generated from wearable devices, home monitoring systems and health records and how they are turned into reports for fitness buffs, homeowners and patients. Think of customer relationship management (CRM) systems, customer surveys, social media posts and review sites and how dashboards are created to help front-line employees perform at their peak. The better grasp of statistics these individuals have, the more insight, value and use they will get from the reports. Giving data-intensive reports to nonstatistical people and expecting them to get the full value of the reports is like speaking English to non-English speakers and expecting them to understand what you’re saying. I’m not saying that everybody needs an in-depth knowledge of statistics, but they could learn some basic statistical concepts to improve how they approach data. Some concepts could include sampling error, the role of biases, hypothesis testing and probabilistic thinking, to name a few.
Statistics and analytics are what help extract value from data. Analytics can help everybody use data to describe the current state, predict future events and make enhanced decisions that capitalize on this knowledge.
To experience the power of advanced analytics and data management to transform your business and deliver unprecedented insights, explore the comprehensive IBM Analytics solution family. For deeper data science insights, peruse Bob Hayes's blogs Getting Insights Using Data Science Skills and the Scientific Method and Data Scientists and the Practice of Data Science.