Business Success Depends On Smart (Not Just Big) Data

Companies are heavily investing in acquiring and developing talent, technology and business processes aimed at collecting and analyzing massive amounts of data, so that they can develop actionable business insights aimed at bolstering customer value. The key driver for digital business transformation is improving the ability to convert data to knowledge and understanding that leads to meaningful and timely action. Yet, a gap exists between what businesses need and what big data technology enables businesses to do, which is still largely at the infrastructure layer where it is stored to be searched and retrieved. In order to move beyond simply collecting data, into a pretty visual story that merely summarizes data, to truly analyzing the data so it can be used to gain the insights that businesses really want, often requires the experience of a data scientist. For marketing organization, a data scientist may be the most important hire.

But what does a data scientist actually do? Who better to ask then Chief Data Scientist for Lithium Technologies, Dr. Michael Wu. Dr. Wu spends his days crunching numbers, testing and building models to try to understand social customer behavior on different social channels to predict customer behavior and its effect on the business. Wu helps us to understand how companies can get the raw data into the information insights that they actually want to make better decisions. Today, marketing and sales organizations are working hard to conquer the data deluge.

Dr. Michael Wu (Twitter: @mich8elwu), Lithium Technologies

5 Steps to Get from Raw Data to Actionable Insights:

1. Start with determining the business problem you're trying to solve - As a starting point in dealing with all of this data, Wu recommends that businesses begin with a business problem. Collecting all the data is important because you don't know what problem you may have in the future, but in order to use the data to take business action, you need to start with a problem. "If you want to have a big data initiative or a big data strategy, first identify some problems so that the data that you collected will have more immediate value," says Dr. Wu. In doing this, data will have more value and a longer shelf life, increasing the likelihood that it will prove valuable in 10 or 20 years down the line.

Once you have collected the data to solve a particular problem, then you have to see what kind of attributes or what information and what insights you can get from this data. Everyone wants to know how to improve their awareness on social media and Wu says that there is lots of data out there about people's consumption of social media or participation in social media. Marketers can actually look at that data and do some simple types of analysis so that they can maximize their social media efforts.

2. Start with descriptive analytics - In order to move raw data to information insights, Dr. Wu says that there are three classes of analytics that people can use. The first-class is called descriptive analytics, which is a summary of historical data that has been collected that is usually shown as a visual dashboard. Dr. Wu says that 80% of most of the analytics that most companies do fall into this category. "You always start with descriptive analytics and then if you get enough data you can become more sophisticated and then you could actually build predictive analytics. And if you are more advanced then you essentially do prescriptive analytics," says Dr. Wu.

There are three general classes of analytics for data reduction and decision support:

1. Descriptive Analytics: Compute descriptive statistics to summarize the data. The majority of social analytics fall in this category.
2. Predictive Analytics: Build a statistical model that uses existing data to predict data that that we don't have. Examples of predictive analytics include trend lines, influence scoring, sentiment analysis, etc.
3. Prescriptive Analytics: Build a prescriptive model that uses not only the existing data, but also the action and feedback data to guide the decision maker to a desired outcome. Because prescriptive models must be actionable and have a feedback data stream, social analytics are rarely prescriptive.

3. Compute sentiment with predictive analytics - The simplest type of predictive analytics that everyone is familiar with is a trend line. You look at the data and then follow some trend and you can see that if you continue to follow this trend tomorrow, or in the future, it will be a certain, predictable value. Dr. Wu says that the interesting point about predictive analytics is that you don't have to just predict the future; you can actually predict things in the past as well. In this case, you are trying to use data that you have to predict data that you don't have. "Predictive analytics is really simple and it's basically what you put into a model and the output of the models tells you something that you don't already know," says Dr. Wu.

In social media, there are a couple of types of predictive analytics that people are familiar with. For example, sentiment analysis is actually predictive analytics. "With sentiment analytics, nobody actually goes out and reports that their sentiment is positive for Apple or Android or whatever. In their natural language, they just say 'I love my iPhone' or 'I love my new android'. Using the natural language as the known data, we build a model that uses linguistic processing, so when people use this type of language; it typically means that they have positive sentiment or negative sentiment. So the sentiment is actually computed and has not actually been measured," says Dr. Wu.

4. Meet KPIs with prescriptive analytics - The simplest example of a type of prescriptive analytics is a Google map; it has prescribed a route for you to get to where you want to go. Like with predictive analytics, once you have a model you can predict things in the past. With prescriptive analytics, can prescribe what you need to do and what you need to focus on in order to get to a particular business key performance indicators (KPI), such as achieving the highest customer satisfaction or the greatest lift in revenue.

5. Get to actionable results - Wu thinks that whether it is descriptive, predictive or prescriptive, the ultimate goal is to help business decision makers to take action on the analytics of their data. "Action-ability is really important. A lot of people say that they provide actionable analytics, but what do they actually mean? Actionable is a type of analytics and is also prescriptive analytics; it tells you a course of action where you can take the action and affect the outcome," says Dr. Wu. If you cannot take an action, then it's not prescriptive analytics.

Dr. Wu explains that with prescriptive analytics there is this notion of what we call a predictive window, which means that within this window the error that you make in prediction is still acceptable. When we talk about action-ability, you have to have another measure called reaction time, which is the time that it takes for you to take action on what you have learned from these predictions. "One of the most important criteria for action-ability is that your reaction time has to be shorter than the predictive window," says Dr. Wu.

Dr. Wu's concludes that, "We have no shortage of data. As big data technologies are commoditized, the accessibility to data will only increase. What we need is smart data analytics to distill petabytes of big data into actionable bits."

You can watch the full interview with Dr. Michael Wu here. Please join me and Michael Krigsman every Friday at 3PM EST as we host CXOTalk - connecting with thought leaders and innovative executives who are pushing the boundaries within their companies and their fields.