Ask an Alum: Ajit Sivadasan

For our Ask An Alum feature, we asked Drucker School of Management alumnus Ajit Sivadasan (MBA, 1999), vice president and general manager for global ecommerce at Lenovo, to share his experiences and insights. Sivadasan is scheduled to serve as a panelist and special guest for Drucker Day 2015 on November 14, 2015.

Culture Transformation is Key to a Data Driven Organization

Much has been written and said about the newfound power of data. In fact, it is hard to get away from the data-talk and the related jargon if you are part of it or immersed in marketing or technology these days. It goes without debate that the fast-paced digital world is in the midst of a dramatic and profound transformation. This is hardly surprising given that an estimated 10 billion mobile devices will be in use by 2020, a billion searches are performed on Google each day, a staggering 294 million e-mails are exchanged daily, and a mind-numbing 50 billion messages are exchanged on WhatsApp every single day. It is safe to assume the changes are irreversible and will inevitably lead to possibilities that will profoundly impact society.

“While new data is available in shiny formats and new jargon, old data suffers from legacy and traditional problems of integrity and compatibility. The cost for cleaning and transforming this can be enormous. Managing and maintaining the ecosystem can be time-consuming and cumbersome, not to mention requiring talent that is sometimes difficult to find.”

There is a certain level of excitement about the possibilities, from new drugs that cure rare diseases to underwater oil exploration to interpreting and analyzing the entire human genome. At the same time there are serious concerns about moral and ethical dilemmas arising from privacy, security, and—in some cases—the implications of artificial intelligence driven by data. The ability to follow, snoop, customize, segment, and target the digital persona has escalated significantly to a point where it is a serious concern for individuals and governments. We have plenty of data being generated and tons of technologies trying to make sense of it. But organizations, for the most part, will face a significant set of challenges before the data become useful for them. For organizations such as governments, it poses a set of deep-rooted moral and ethical challenges. In any case, button down, strap up, and get ready for the ride.

Companies today face significant challenges before they are going to be able to take advantage of this paradigm shift. Several companies have ambitious plans (23 percent of companies polled) to roll out Big Data initiatives to drive internal productivity, external sales, and marketing activities. However, the successes of these initiatives have been all over the place. Some companies that deal with data and have done so for years have made some progress, but still not at the hyped-up level made out by those that sell these solutions. While new data is available in shiny formats and new jargon, old data suffers from legacy and traditional problems of integrity and compatibility. The cost for cleaning and transforming this can be enormous. Managing and maintaining the ecosystem can be time-consuming and cumbersome, not to mention requiring talent that is sometimes difficult to find. If you are an organization that wants to transform itself into a thoroughly data-driven organization, there are more things beyond technology that have to change fundamentally. Such changes are at an organization’s core and present more difficult management challenges than most people can overlook. Here are some observations, problems we have seen, common mistakes people make, solutions, and recommendations if you are embarking on a data journey to transform your business.

Data is Never Clean
Operate under the assumption that data is never really clean. It requires people, processes, and tons of energy to visualize and get into formats that are ultimately useful. The process is continuous and will continue to be challenged by evolving technology, business requirements, and the very dynamic customer landscapes the business intending to deal with it finds itself in. Data comes from multiple sources: transaction processing, click stream, segmentation, medical records—you name it—and naturally it suffers from compatibility issues. So if we have to make sense of stuff, we need to be able to handle multiple types of data in a flexible way, make it clean, and be certain it works nicely with everything in your system. We need to be realistic about this and build systems that allow for backward and forward integration of technology so we don’t end up with monolithic solutions that later become boat anchors.

Message: Be ready to go through a painful process of inventorying, ETL-ing (Extract, Transform, and Load), and integrating data and sources as you begin to think about making sense of the various disparate systems.

Technology is Important, But Really Not a Deal Breaker
Move under the very real assumption that technology will evolve rapidly. We will see lots of innovation on data management solutions driven by venture capital and social funding. While you don’t want to move from one shiny object to another, you also should not lock yourself into technology that is old and clunky if you have access to new technology that makes your life easier. So develop and use frameworks that are built with flexibility in mind, preferably with the explicit mandate of using a building-blocks approach.

Message: Think clearly through the various levels and layers of data and technology. Design for flexibility. Don’t put all eggs in the same basket. Ensure individual pieces aren’t too big to fail.

“In many ways, organizational culture drives your data strategy success and its adoption even more than your choice of technology.”

Use Culture as a Pivot For Change
In many ways, organizational culture drives your data strategy success and its adoption even more than your choice of technology. My own experience is that it takes upwards of three to five years to effect something reasonably good. Here is the challenge: Moving from data reporting and basic insights to driving significant and deep-rooted insights in a continuous manner. Many individuals and leaders are skeptical because their experiences have been painful, inconsistent, and prone to wide fluctuations in the quality of insights. Therefore, trying to implement an organization-wide transformation requires fundamental cultural transformation. This is time-consuming, challenging, and expensive.

Message: Take it seriously and involve everyone; the skeptics and the evangelists. Make it visible and use experts to make it your company’s core competency.

Precision is Overrated—Expect to Fail Multiple Times
The whole data transformation exercise can be incredibly frustrating to both the ones developing the systems as well as the ones using them. The funny thing, however, is that these systems should be developed precisely to address that aspect. They should be designed and built for people to experiment and try different things. The more there is room to experiment, the better the adoption and, ultimately, the quality of output. So it is not uncommon to have several projects in flight and several experiencing multiple failures, from small ones to really big ones. A small one can be a data compatibility issue between two systems that pushes timelines. A more serious one is investing significant time into an initiative only to realize that you cannot get to where you want to go and must reset and restart. I have been part of both and it has helped me learn and avoid potentially costlier mistakes down the line. As a leader, you have to change your mindset on what constitutes “success.” Rigid, black-and-white measurement processes may not yield the kind of flexibility needed to constantly tweak and improve. Again, if culture is a pivot, then this simply becomes an extension of that thought process.

Message: Empower your people to be able to play around with stuff. Traditional measures of success may be limiting. New measures such as “time to value” or other metrics may be better suited.

In conclusion, technology may not be the biggest obstacle for organizations as they try to transform themselves into level 5 data organizations; the company’s culture might be. It is important to understand, articulate, and execute on a deliberate strategy that looks to fundamentally change the company’s thinking and culture about data.