Eliminating Data Silos for Unified Workflows

John Williams, Executive Director Of Enterprise Data And Advanced Analytics, Racetrac

John Williams is a pragmatic Enterprise IT leader with an excellent track record of architecting and implementing complex data solutions across diverse industries. Having started his career at FedEx as a programmer, he went on to work at multiple companies as a data warehouse developer and data architect. Presently serving as the Executive Director of Enterprise Data and Advanced Analytics at RaceTrac, he is squarely focused on promoting a progressive, data-driven culture.

In an interview with APAC CIOoutlook, John Williams sheds light on some of the current challenges prevailing in enterprise data management and how companies can break the barriers of data silos to boost productivity, drive down costs, and facilitate informed decision-making.

Can you walk us through your roles and responsibilities when it comes to enterprise data management?

Blinders help horses stay focused during races and not encounter any distractions. it is the exact opposite of what we want with data. we want to see all possibilities with data collected from different sources to get a holistic, 360-degree representation to save valuable resources

My primary responsibility revolves around data. It mainly entails pulling data from various applications and systems and amalgamating it to accelerate decision-making. The next area of focus is business intelligence, where reports, dashboards, and analytics are taken into account. We are also working on a new data governance program for effective data management.

What impact do data silos have on overall enterprise operations and productivity?

Data silos do not provide you with a single version of data management, which impacts productivity, customer experiences, and costs and lead to an overall lack of accountability. In my opinion, it is much like horse racing.

Blinders help horses stay focused during races and not encounter any distractions. It is the exact opposite of what we want with data. We want to see all possibilities with data collected from different sources to get a holistic, 360-degree representation to save valuable resources.

Let’s say you have a marketing department rolling out a product in a nationwide campaign in Florida, and the stint became a massive success, significantly ramping up sales. From one perspective, it might seem that the campaign is the major driver of sales. But from a different perspective, it might be a hurricane in Florida that caused the sale to spike. So it turned out that the weather was the prime factor that spiked the sales rather than the marketing campaign. This would lead the decision-makers to redirect the advertising budget for subsequent events.

At RaceTrac, we have successfully eradicated silos created due to data flowing in from three different systems––the fuel pump, point of sale system, and underground system. We identified that the data was not in sync, which led us to understand that there was potential fuel fraud occurring. This discovery helped us save approximately $30 million in a year. We also leverage predictive analytics to resolve issues on time, which improves productivity and efficiency.

Setting data policies on all kinds of data and their access comes under data governance. Breaking down data silos and adding data governance ensures complete and accurate data from a single source of truth, improving data security and transparency.

How has the enterprise world evolved to combat these challenges with technology and innovation?

Enterprises tend to be slow to change. Technology, on the other hand, evolves at a pace much ahead of enterprises. The reason that is paramount for the slow transitions in enterprises is that organizations make considerable investments in technologies, hoping for drastic returns on investments. They are often skeptical about change. However, data, touted as the new oil, has made them realize that time is of the essence. Migrating to the latest technologies can enable them to make better and faster decisions and improve the bottom lines.

Many a time, decision-makers are perplexed by the numerous technologies in the market, which deter them from taking the leap. Things further get complicated when you need an in-house team of technology experts and approval from the management team to invest in the right technology.

What leadership strategies do you implement for guiding and leading a team toward successful project execution?

For successful project execution, we implement an agile methodology for collaboration with stakeholders and continuous improvement at every stage. Business stakeholders must be made part of the process for thorough understanding and transparency. I believe in the philosophy; “we all win and lose together.” Every person involved has a role to play, and communication is the key to eliminating any barriers between business stakeholders and other teams.

Another aspect is intensive training sessions for all the business pillars––enterprise data, business intelligence, and data governance. We have training programs for business intelligence where the six sigma approach is followed with certification in white, green, and black belts. Each focuses on increasing the level of expertise in Power BI and also managing a project in different stages.

How do you envision the future of enterprise data management? According to you, how can breaking down the existing data silos create a unified workflow and operational synergy?

The real change would materialize once enterprises understand the real value of data assets. A company can significantly improve its bottom line by leveraging the data they have in-house and using analysis reports. Once companies understand that, they will undeniably move forward with data.

Getting into predictive and prescriptive analyses creates a real impact on an organization’s bottom line and revenue stream. With predictive analysis, companies can understand customer behavior based on past data, while prescriptive analysis helps in devising a course of action to follow based on the data. Understanding customer purchase patterns and creating customer loyalty would amplify revenue generation. Once enterprises realize how to use the data to create a revenue stream, data silos will cease to exist, and you will have unified workflows and data strategies.