Navin Advani, VP of Enterprise Data Management at global food distributor Sysco, presented during re:Invent and focused on the people, process, and technology aspects of deploying successful analytics at enterprise scale. As Eric Ries, says in the seminal book “The Lean Startup”, the oft forgotten management elements frequently dismissed as the “boring stuff” are what makes cool technology tick – and allows that technology to be adopted widely. And it was fascinating to watch Navin distill the key motivators, business drivers, delivery models, change management, org structures, governance, training, and the process improvements that underpinned their analytics strategy.
Starting with the business drivers, Sysco’s analytics charter was informed by the goals of its three-year operating plan: Unlock savings, drive top line growth and market share, enabled by quick actionable insights. Navin highlighted that Sysco’s current state of analytics infrastructure was not fit to deliver that plan: It was plagued by the lack of abilities to analyze large volumes, reporting inconsistencies, long lead times, creeping cost of ownership and scalability, and stability issues. Navin and team needed an enterprise view of its customers, products, and suppliers that would deliver more intuitive information to internal and external customers through a self-service reporting model, which in turn would ultimately enable revenue growth and operational efficiencies with better data management and better decisions.
Sysco undertook an enterprise-wide transformation exercise and their data and analytics processes morphed and matured throughout the years. In certain business functions, there has been a shift from consuming hindsight information using row and column-based flat reports to ad-hoc data analysis for deriving insights to support better decisions – and even developing some predictive analytics. This has provided a launch pad for further development of Sysco’s analytics capabilities, driving them to a data driven culture, and serving as a catalyst for the business intelligence roadmap.
With “Tableau on AWS”, Sysco is able to implement their vision for a “Sysco Ecosystem for Enterprise Data” aka “SEED”. Navin talked about how “SEED” was envisioned to scale with evolving business needs and provided a foundation for data governance and data security. SEED was cloud native. This meant that Sysco could move from “build first and see later” to a more agile and “continuous improvement” approach towards analyzing business data that involved setting up the ability to quickly stand up sandbox environments for experimentation. “SEED” was also developed with broad and deep functionality to support a range of use cases across multiple functions including marketing (Churn Analysis, Share of Wallet, ML for future promotions etc.), Merchandising (Assortment optimization, Market Basket, Lotting using decision trees etc.) and Revenue management (Predictive pricing simulations, Pass thru predictive pricing, Margins review etc.)
The agility, governance, cost leverage and discipline that SEED has brought to how data is mined and used at SYSCO has delivered amazing results. People feel empowered to use data in their day to day work and usage numbers have exploded from 64 to 12,713 (roughly a 20x increase) in three years.
You can watch the Sysco talk in full here.