In the new movie "The Imitation Game," Benedict Cumberbatch portrays Alan Turing as he develops the first computer to decode encrypted Nazi messages. The emotional backstory aside, the importance of Turing’s story came from the genius way in which he dealt with data sets with combinations in the millions of millions of millions. Whenever working with numbers that have 18 zeros behind them, the functionality and significance can be obscured be the sheer size of the data. What comes out of these giant data collections can change the way we do business, assuming that we know what we are analyzing.
Redefine Knowledge
Editor-in-chief of Wired Chris Anderson tells use to throw out all of the other sciences when it comes to big data. We do not care why customers act the way they do. We just need to know that they act. This is the offering that large data mining gives us. When we create our customer relationship management platforms, we need to define our customer goals, called drive factors. When using customer relationship management systems, the new paradigm for knowledge dictates that we only focus on the consumer’s drive factors and not on the things that influence the factor. In other words, if 90 percent of your customers like the outcome, there is no need to figure out why they like it. Big data tells us to keep doing more of what works and analyze the remainder for new trends. It also says that we redefine analytical outcomes by the way we develop them and not by their final results since these can change. This is where experimentation becomes a business’s greatest tool.
Experiment Then Experiment Some More
Imagine a spider web with millions upon millions of threads, each interwoven into a complex structure. Pluck one and the remainder vibrates in response. Large data sets are the filaments of this intricate nodal network of information. Savvy companies use simple and frequent experimentation to find out what variables change in response.
Co-founder of Capital One Nigel Morris says that their IT and marketing team conduct around 65,000 tests each year. Most of these are not grand scale experiments. Instead they are small tweaks that reflect on the outcome of the consumer’s transactional experience. These experiments may include such minor customer care adjustments as the use of a first name and its relationship to overall satisfaction. The statistical analysis of these experimentations becomes somewhat significant and should include structural equation modeling which adds a chronological component to your mathematics. Not only are you creating minor changes with potentially substantial results but you are placing them into an order that offers the greatest customer care experiences.
IRB For Big Data
Much like Pandora’s Box, just because the information is in front of you does not mean that you should look at it. Educational institutions require approval from an Institutional Review Board before scientists can perform experiments that involve living subjects. This includes surveys and observational analysis. Big data investigation does not require this, leading to ethical and public relations nightmares.
Recently research was released that used proprietary data from Facebook. The backlash was huge with claims of ethical and privacy violations. Even though the research was valid and sociologically very useful, Facebook took a hit to its reputation. When doing your data analysis, be certain to weigh the consequences against public perception. Positive consumer experience is subjective so be sure to understand what beliefs your customers may make from your data outcomes and methodology of analysis.
Customer Service and the Toaster
In 1999, IT executive Kevin Ashton first coined the term "Internet of Things" (IoT) to describe the network of devices that we are seeing interconnected throughout the world. IoT and big data have opened up a new world to customer care in which people are not the prime recipients.
Assuming permission, a customer care team can automatically upgrade software on functional devices like toasters, televisions, or automobiles. They can also access informational data, gaining insight into the daily usage of common appliances. From there it becomes a function of marketing, letting potential customers know that their equipment is being made better even when they do not know it.