What is data analytics and how it is used by large organizations to support strategic and organizational decisions? Senior leaders offer insight into the problems and opportunities involved.
Most data and analytics (D&A) conversations begin by focusing on technology. It is critically important to have the right resources, but executives too often ignore or underestimate the value of the people and organizational components needed to create a productive D&A process.
When that happens, D&A initiatives that fail — not deliver the insights required to move the organization forward, or inspire trust in the actions necessary to do so. The stakes are high with International Data Corporation predicting global D&A market spending to surpass $200 billion a year by 2020.
A stable, efficient D&A feature encompasses more than just a stack of technology, or a couple of people isolated on one level. D&A will be the organization’s heartbeat, integrated into all leading sales, marketing, supply chain, customer service, and other core functions decisions.
Why can I develop successful D&A capabilities? Start by creating an enterprise-wide plan that provides a clear picture of what you’re trying to achieve and how progress will be evaluated.
One of America’s prominent sports leagues is a perfect example of an organization making the most of its D&A feature, applying it to cost management plans, for example, reducing the need for teams to fly back-to-back nights from city to city for games. Throughout the 2016–2017 season, thousands of travel-related restrictions, player exhaustion, ticket sales, arena capacity, and three major TV networks had to be taken into account. With 30 teams and 1,230 regular season games stretching from October through May, there were trillions of scheduling choices available.
Companies should follow the lead of the league by understanding first that good D&A starts at the top. Make sure the leadership teams are entirely engaged in the company in identifying and setting goals. Avoid allowing the setting of objectives and decision-making to take place in organizational silos that can generate shadow technologies, conflicting versions of the reality, and paralysis of data analysis. Ask: Is the aim to help boost the company output before launching some new data analysis initiative? System jump start and cost efficiency? Drive policy, and speed up change? Growing market share? More successful innovation? Any of that?
Leadership teams must understand that it takes bravery to be successful because, as they embark on the journey, data analytics observations will always point to the need for decisions that may entail a course correction. The leaders need to be frank about their ability to integrate the findings into their decision-making. Further, they should keep themselves and their teams accountable for that.
Consider a large global life sciences company that spent a huge amount of money to develop an advanced analytics platform without knowing what it was supposed to do. Executives allowed their development team to buy a lot of items, but none understood what the developed tools were meant to do or how to use them. Luckily, before it was too late, executives identified the issue, undertaking a company-wide needs assessment and restoring the platform in a manner that inspired trust in its ability to drive productivity and promote business transformation.
In another instance, a global financial services company, focused on stakeholder expectations, has developed a robust development infrastructure. But executives soon discovered after creating it, that they lacked the organizational structure and resources to use the platform effectively. When these requirements were met, the organization was able to use a great platform to generate substantial operating cost savings.
Data analytics is the most in-demand technology skill for the second year running, according to KPMG’s 2016 CIO Survey. Still, almost 40 percent of IT leaders claim they suffer from skill shortages in this critical sector.
Formal, organized structures, procedures, and people committed to D&A can be a competitive advantage, but this significant opportunity is lacking in many organizations. Companies who develop a D&A infrastructure to meet their business needs have in our experience teams of data and software developers who are experienced in using big data and data scientists who are entirely focused on a D&A initiative.
Although processes vary, the team should be integrated seamlessly with existing D&A suppliers and customers in the sector, working in collaboration with non-D&A colleagues — people who understand both the market problems and how the business analytics functions — to set and work towards practical and specific strategic objectives. The teams will need full executive leadership support, and their priorities should be aligned entirely with the company plan.
In an era in which data is generated on a scale well beyond the capacity of the human mind to process it, business analytics leaders need D&A that they can trust to inform their most important decisions— not just to cut costs but also to achieve growth. And the best would use D&A to predict what they want or need from their customers before they even know they want or need it.
Volatility, sophistication, and confusion can better characterize today’s business analytics decisions governing the macroclimate. In this uncertain climate, forward-thinking companies are identifying and exploiting data as a strategic tool to improve their competitive edge. Data Analytics facilitates proactive decision-making by offering data-driven insights into products, consumers, competitors, and any aspect of the market climate.
Today analytics are applied on a need-based basis in most organizations. Although most companies are still considering making investments in data analytics and business intelligence, they need to realize that the process of incorporating advanced analytics into the corporate structure requires far more than investing in the right people and resources.
A data-driven culture is the core of this framework and is a crucial factor for the effective introduction of analytics into the organizational system. The integration cycle begins with a data-driven resolve. Big data analysis and advanced analytics must be accepted at the corporate level as an operational feature to be powered by the data. Projects and assignments undertaken must be analyzed from an analytical perspective, in which the quantity, consistency, consequences of the produced data, and how to use the data at hand are of primary importance.
Few deny that more data is available to organizations than ever. But it’s easier to say than done to derive meaningful insights from that data— and convert knowledge into action. We talked to six senior leaders from major organizations and asked them about the challenges and opportunities involved in the adoption of advanced analytics: Murli Buluswar, AIG’s Chief Technical Officer; Vince Campisi, GE Software’s Chief Information Officer; Ash Gupta, American Express ‘ Chief Risk Officer; Zoher Karu, eBay’s Vice President of Global Customer Management and Data. It is an edited transcript of their remarks.
Here are some professionals talking about their experience with the integration of data analytics at their organizations, proving how data analytics is helping their growth.
Murli Buluswar, Chief Technology Officer, AIG: The biggest challenge of making the transition from an information culture to a learning culture— from a culture that primarily relies on heuristics in decision making to a culture that is much more analytical and data-based and accepts the power of data and technology — is not the cost. Initially, it ends up being just imagination and inertia.
What I’ve discovered in my last few years is that the power of fear to grow to think and act differently today is immense, and to ask questions today that we weren’t thinking about our positions beforehand. And it’s that shift in mindset — from an expert-based mindset to one that is much more fluid and learning-oriented, as compared to a fixed mindset— that I think is essential to every company’s sustainable health, big, small or medium.
Ruben Sigala, Chief Analytics Officer, Caesars Entertainment: What we found challenging, and what I find challenging in my discussions with many of my peers, is finding the collection of tools that enable organizations to generate value efficiently through the process. I hear of individual wins in some projects, but creating a more unified environment where it’s completely incorporated is something I think we’re all grappling with, partially because it’s still very early days. We have been thinking about it in the past few years quite a bit. The technology is always changing. Also, the sources are still evolving.
Zoher Karu, Vice President, Global Customer Optimization and Software, eBay: Data protection is one of the most significant problems and what’s shared and what isn’t shared. And my view on that is customers willing to share if it returns interest. Single way sharing will no longer float. So, how do we secure this knowledge? How do we leverage it and become a partner for our customers rather than just a seller for them?
Ruben Sigala: You have to continue with the organization’s charter. You must be very precise about the nature of the role within the company and how it is expected to communicate with the broader enterprise. Some organizations begin with a relatively oriented view of supporting conventional functions such as marketing, pricing, and other specific areas. And then there are other organizations which take a much more comprehensive view of the market. Also, the elements need to be defined first.
That helps better inform the correct structure, the forums, and finally, it sets the more granular operational levels such as training, recruiting, and so on. It is essential to align yourself with how you are going to run the company and how you are going to communicate with the larger organization. And there it will fall in line with everything else. It is how we set out on our journey.
Vince Campisi, Chief Information Officer, GE Software: One of the things we found was when we began and concentrated on a goal, it was a perfect way to generate value and get people excited about the opportunity rapidly. And it took us to places we didn’t plan to be going before. Then we can go after a given outcome and seek to coordinate a collection of data to achieve that outcome. When you do so, people start bringing in other data sources and items they want to link. So it’s just getting you to a point where you’re going after the next outcome you didn’t foresee going after before.
You’ve got to be able to be a little flexible and versatile about how you think things. But if you start with and deliver one result, you’ll be shocked as to where it takes you next.
Ash Gupta, Chief Risk Officer, American Express: The first change we had to make was just to increase the quality of our results. We’ve got a lot of data, so sometimes we just didn’t use the data, so we didn’t pay as much attention to its accuracy as we do now. It was one, to ensure the data had the correct history, that the data had the right intent to represent the customers. It is a road, in my opinion. We have made strong progress, and we expect this progress to continue in our program.
The second field is working with our people and making sure we centralize all aspects of our business analytics. We centralize our resources, and we democratize their use. The other thing, I believe, is that we understand as a team and as an organization that we do not have enough resources ourselves, so we need cooperation from all kinds of organizations beyond American Express. This partnership comes from innovators in technology, and it comes from data providers and analytical firms.
We need to put together a full package for our business colleagues and partners so that it is a compelling case that we are collectively improving stuff, that we are co-learning, and that we are building on each other.
Victor Nilson, senior vice president, big data, AT&T: The consumer experience also begins with us. That’s what matters the most. We now have a growing range of very complex items at our customer service centers. Only the simple products often have potentially very complicated problems or solutions, because the workflow is very complicated. And, how can we improve the process at the same time for both the customer-care agent and the client, if there is an interaction?
We’ve used big data strategies to evaluate all the various permutations and improve the knowledge and solve or strengthen a specific situation more quickly. We take out the uncertainty and make it a smooth and actionable task. At the same time, we should evaluate the data and then go back and say that, in a particular case, whether we are proactive or not for optimizing the network. So, not just for customer service, but also for the network, we take the optimization and then tie it together.
Vince Campisi: I’ll give you a personal viewpoint and an outer viewpoint. One is that we do a lot of what we call creating a digital thread— how you can link innovation to a product through engineering, manufacturing, and out. And that’s where we focus on the brilliant factory.
Take the driving optimization of the supply chain as an example. We’ve been able to take over 60 separate silos of direct-material sourcing knowledge, exploit analytics to look at new relationships, and use machine learning to find enormous amounts of flexibility in how we source direct materials that go into our product.
An external example is how we use analytics to enhance the efficiency of properties fully. We call it the administration of asset results. And we are beginning to allow digital industries, like a digital wind farm, where analytics can be leveraged to help the machines optimize themselves. And, you can help a power-generating company use the same wind that comes through, and by making the turbines pitch themselves correctly and knowing how they can maximize the wind speed. We have also shown the potential to generate up to 10 percent more energy out of the same amount of wind. It’s an example of using analytics to help a company achieve higher yield and productivity.
Ruben Sigala: There’s intense competition for intellectual talent. And it’s challenging to retain and sustain a talent base within an organization, mainly when you see that as a core competency. What we’ve focused on is building a forum that appeals to what we think is a value proposition that is important to individuals looking to start a career or carry on a career in this area.
When we speak about the value proposition, we use words such as getting a chance to influence the market results, to provide a wide variety of analytical activities that you will be faced with regularly. But, generally speaking, being part of an enterprise that sees this as a vital part of how it competes in the marketplace— and then consistently operating against that. You have to have strong training systems in part and to do it well, and you have to have obvious ways of contact with the senior team. So you must always be a member of the team that is implementing the company’s plan.
Murli Buluswar: I found that concentrating on the basics of why science was developed, what our goals are, and how being part of this team would shape the team members ‘ professional evolution was very insightful in attracting the talent we care. And then, of course, there comes the much tougher aspect of living that promises day-in, day-out.
Yeah, it is necessary to have the capital. My money theory is that I want to be within the 75th percentile range; I don’t want to be in the 99th percentile. Because irrespective of where you are, some people— especially people in the data-science function— have the chance to get their pay increased by at least 20 to 30 percent if they wanted to transfer? My aim is not to seek to cut the distance. My goal is to build an atmosphere and a community in which they see that they are learning; they know that they are working on issues that have a broader effect on the business, industry and, through that, on society; and they are part of a dynamic team motivated by why it exists and how it defines success.
Focusing on that, to me, is an utterly vital enabler to attract the caliber of talent I need, and everybody else would.
Victor Nilson: Talent is all about that. You need to have the data, and, naturally, AT&T has a wealth of data. But this is meaningless, without talent. The differentiator is talent. The best talent will go and discover the best technologies; the right talent will go out there and solve problems.
We also helped, in part, contribute to the growth of many of the innovative innovations emerging in the open-source culture. We have the tradition of sophisticated lab methods, and we have the Silicon Valley that is evolving.
But we do have mainstream talent around the world, where we have very advanced engineers, we have all-level managers, and we want to grow their abilities further.
So just this year alone, we have delivered over 50,000 Big Data related training courses. And we continue to make progress on that. It is just a continuum. It could be either a one-week boot camp or advanced data science at the PhD-level. But for those who have the aptitude and interest in it, we want to continue cultivating the talent. We want to make sure that they can improve their skills and then tie it to enhance their output along with the instruments.
Zoher Karu: Talent is crucial in every path through data and analytics. And analytics talent alone, in my opinion, is no longer enough. We can’t have singularly competent men. So the way I develop my organization, I’m searching for people with a minor and a major. You may be significant in analytics, but in marketing strategy, you may be minor. Even if you don’t have a child, how can you interact with other parts of the organization?
Otherwise, the mere data scientist won’t be able to speak to the database administrator, who won’t be able to talk to the market analysis guy, for example, who won’t be able to speak to the owner of the email channel.
Data Engineering
AI & Machine Learning
By Use Cases
Business Intelligence & Tableau
1 Comment