Connect data teams, tech teams, and business decision-makers to unleash data value
Venkat Raghavan’s role as Global Head of Enterprise Analytics, Tesco, is right in the thick of this revolution. Three years into a career at the UK-based retailer, he is setting up an enterprise-wide center of excellence for the £50 billion business as part of Tesco Business Services in Bengaluru, India.
With a career primarily focused on customer analytics and stints at Walgreens, Walmart, and BCG GAMMA, Venkat finds himself at a business with a long-established lead in customer data. Tesco established the Clubcard loyalty scheme with Dunnhumby back in 1995.
Clubcard has more than 20 million users in the UK alone. The data it provides means Tesco has been able to turn its back on the kind of data integration between third-party sources delivered by the growing array of customer data platform providers discussed in the HFS Point of View: Markets need customer data platforms to accelerate to One Office success.
Data teams can help the business understand customer behavior at the individual level—at least online—and by aggregating anonymized datasets at the cohort level, too. That understanding can be applied to both digital and brick-and-mortar domains. But customer data is only one facet of a business’ cross-functional data flow that Venkat says is rapidly becoming essential for success.
Data leaders must think of themselves less as the people who say “no” and more as the people who enable the flow of data to the right person (or system), for the right reason, at the right time, a goal very much in line with the HFS view of data strategy.
As HFS CEO and Chief Analyst Phil Fersht puts it, “Data drives everything. It is your strategy. Your core processes must be redesigned to get this data you have to have. They must be automated and run in the cloud. Then you can apply many flavors of AI to get data at a speed and precision you never thought possible.”
Source: HFS Research, 2022
Enabling data flow means balancing several conflicting needs. In some cases, the drive toward increasing the speed and democratization of data access conflicts with privacy and security needs.
As data needs increase, so does the complexity of judging data’s value. Who should be allowed access to what? How can you enable the right usage of the data? These are the challenges for data governance. Tesco is developing the right architecture to respond because governance may be all about enabling speed in some cases but protection in others. Far from being a policing function, data governance must shift focus to enabling value creation.
Venkat recalls how when he started his career 20 years ago, functional teams would focus on functional problems. The customer team wanted customer data, the finance team wanted finance data, and so on. The rise of data science brings positive disruption to this requirement. Data scientists need multiple variables to create prediction models. They need a customer view, an inventory view, a vendor view, and even a view of competitor activity.
This demand from the data scientists increases the value of, and priority for, cross-functional, horizontal datasets organized around a data problem’s solution rather than locked into traditional functional data silos. Venkat says organizations must be culturally ready for this change, requiring a OneOffice™ mindset to master.
Databases will have to provide multi-layer approaches to governing data, depending on who needs access and to what end. The data scientist needs cross-functional and deep data access. Local decision makers may need fast and easy access. Executives need protected and secure access. Venkat argues that governance is now less about rigid rules and control and more about adapting to different needs with agility.
To support this agility, data team organizations, even in large enterprises, must replicate the flexible and problem-focused “squads” familiar to startups, says Venkat.
Venkat says there is often a disconnect between how systems are architected and data requirements. He gives the example of self-serve checkouts. Not only should we see these as a tool through which someone will shop, but they should also be considered a very important tool to capture data. We should not have to send someone out with a clipboard to observe behavior and ask questions. If we design data collection into the customer journey, we can learn what the customer likes and does not like, what takes them more time, and what makes them give up—and feed this back into improvement cycles.
He says data teams should spend time with both decision makers and tech teams to think through data flow from system to analytics to decision-maker… and the return path, whether the system is customer, partner, or robot facing. For example, Tesco now has nearly wholly robotic fulfillment centers.
Data science, multi-layer databases, the increased use of AI and ML, and the application of agile ways of working are valuable to leaders because they facilitate smart responses to rapid change. We are living through an era of rapid and regular context shocks—Brexit, COVID-19, inflation, Ukraine—they keep coming, thick and fast. Adopting an agile approach to data governance turns the discipline from constraint to enabler. Leaders must stop seeing data governance as a necessary evil or unavoidable cost and instead consider it a value center supporting the insights and predictions to maintain a competitive edge.
Register now for immediate access of HFS' research, data and forward looking trends.
Get StartedIf you don't have an account, Register here |
Register now for immediate access of HFS' research, data and forward looking trends.
Get Started
