Contact us
We are taking you to another website now.

Analytics: A Still-Emerging Business Frontier Interview with Prabhudev Konana

Prabhudev Konana
Associate Dean of Instructional Innovation Professor at UT-Austin

In the age of big data, companies have unprecedented access to information on everything from connected IoT products to customer opinions. But collecting large amounts of data and wringing business value from it are two different things—and most companies struggle greatly with the latter.

What is holding companies back from achieving greater ROI? Which aspects of analytics are important looking forward? Which are over-hyped?

To delve into these questions, TCS spoke with Prabhudev Konana, the William H. Seay Centennial Professor of Information Management, and past chairman of the Department of Information, Risk, and Operations Management in the McCombs School of Business at University of Texas at Austin, where he founded the Master of Science in Business Analytics degree program. 

The real analytics struggle for many companies starts at the beginning, says Konana. “Knowing the right business questions to ask as you analyze data is vital. If you do not ask the right questions, you will never realize significant ROI.”

TCS: How do forward-looking companies measure the success of analytics projects and which emerging metrics might be important in a few years time?

Prabhudev Konana: The most important business metric is the basic one: dollars and cents. Many companies are using social media analytics to target the right customers, but at the end of the day, are they converting those potential customers into real customers? That is the biggest missing piece in analytics today: how it affects top-line and bottom-line growth.

Once you answer that question, you can drill down to the more operational level and choose the right metrics to support that growth. For example, if you are dealing with credit card fraud detection, your metrics can gauge whether you are reducing fraud situations or flagging incidents before they happen. If you are looking at security issues, your metrics should be able to detect intrusions and spot potential security violations early.

However, the benefits eventually become elusive when these metrics inevitably become widely used and part of the status quo. For example, competitors might invest in metrics that measure customer retention for the same customer base, so every time that you improve, somebody else improves too. So it has less and less of an impact on the bottom line. But it will hurt you if you do not make the investment in analytics. Companies that have superior analytics capabilities in the marketplace will benefit. Others may perennially play a catch-up game.

TCS: Today companies can tap into a range of data sources—social media, mobile apps, geolocation data, government databases, and news feeds. Which emerging data sources intrigue you?

Konana: The biggest source is going to be the Internet of Things, which is making it possible to collect massive amounts of data, including mobile data from smart cars and driverless cars. Data from the IoT is creating smart homes and we are moving towards smart cities. Healthcare is going to be massive in this area; we will see lots of new interesting data sources coming in.

My problem is that over 90% of companies still have no idea what to do with social media data, mobile data, location data, and so on. We are still learning. IoT is the next big thing, but some companies are still struggling with what to do with Facebook data. We are still trying to master the existing data sources to create value.

TCS: Many business users are frustrated with today’s data analysis dashboards, because they are not terribly usable. Do you have any thoughts on what is coming and if that situation will improve any time soon?

Konana: I think it is going to get better. The old dashboards had problems because they were not fully integrated and companies had to force-fit the information sources. Eventually we are going to see ‘micro-level’ dashboards that are more meaningful; they will show things such as “this is our business target; this is what you will reach; this is how we can get there; this is a shortfall.” You can see these kinds of dashboards appearing at the business unit level and at the corporate level.

Over 90% of companies still have no idea what to do with social media data, mobile data, location data, and so on.

The behavioral side must be managed, though. Dashboards can backfire, because if everything goes well, people may feel complacent, and when things are not going well, they may panic. Having too much information on the dashboard can hurt the morale of the company and the way people work. Nonetheless, dashboards can be powerful.

TCS: Can you share an example of a company that you think is doing a better than average job of getting ROI from its analytics?

Konana: Very few companies come anywhere close to the Amazons and Starbucks of the world, the new-generation companies. Traditional large companies have a tougher time, especially if they need to integrate their analytics for online and offline sales. Comparing internet years and the traditional years is like comparing human years and dog years. A traditional company can spend seven years to develop the degree of integration that a new-generation company can do in one year.

That being said, analytics is still a strategic necessity for any kind of company today, and if you do not do it, you are finished. Best Buy has done an extremely good job in using analytics. While we all focus on the problem of ‘showrooming’ (visiting the Best Buy store and then scouting for the best deal online), Best Buy is investing lots of resources to understand what people are saying, what they might buy, and how they might be influenced. They are using Facebook and other social media very well. Home Depot is another example.

TCS: Visualization technology looks to be an important trend for companies seeking better ways to present data to business users. What do you think?

Konana: Visualization is important, but it is really the last stage of the analytics process, when you are presenting insights to senior management. Before that can happen, people must define what they want from the analytics. A lot of senior managers say that their industry is moving fast and there is lot of hype about visualization, so it is something the company should be doing. However, when you drill down, many companies still do not know what questions to ask.

Along with not being clear on what visualization and analytics should accomplish, senior managers are often disconnected from the talent in the company who can do the actual dirty work of analysis and bring it to them. This is a classic business/IT alignment problem.

TCS: Companies are fighting for scarce analytics talent. How important will it be for companies to train up analytics expertise in-house, in addition to hiring outside analytics talent?

Konana: Analytics by definition is very technical and quantitative, and it gets harder to teach the longer a person has been away from mathematics. Many big-name companies found this out when they tried to bring in people with four or five years experience and make them into analytics specialists. Many of them had forgotten their math because they had not done it for years.

Bringing in people with math skills intact, but with less business experience, brings a different set of problems: They often do not fully grasp the complexities of what needs to be analyzed. Those who understand the math often can not explain how it relates to their company’s or customers’ goals. Sometimes that means completely detaching from all the complexities of the predictive models and explaining something simply. That is not easy to do.

That is why the University of Texas requires students to learn accounting and finance along with analytics. We do this because even though they might do social network analysis, they must be able to relate all these metrics to business growth. These metrics are the ones that count.

TCS: Which aspects of the future of analytics are perhaps overhyped?

Konana: The use of social media is one. Companies have been bragging about using social media, surfing data, and analyzing location data. These are noisy, messy data sets and you need specialized people to be able to do it. I have a feeling that many companies will fail trying to create those capabilities inside the firm.

The real hype is that somehow analytics is going to magically change your business. You still need a well defined strategy, and you have got to use data in order to push your strategy ahead.

TCS: What excites you about the future of analytics?

Konana: Healthcare. Hospitals are trying to analyze data so they can understand for example, drug interactions, or early symptoms. According to a recent New York Times article, Microsoft researchers were able to identify symptoms for cancer by analyzing an individual’s web searches. This could quickly identify people at risk for cancer. The challenge will be embedding healthcare analytics capabilities into tools where people do not really have to understand programming to use the data productively.

Another exciting tool is sentiment extraction; for example, what are people saying about my product? Products are being developed that help companies figure this out without doing massive programming; you just feed the data and the technology has the capabilities to do it.

About the author(s)

Thank you for downloading