Data science company best known for retail data analysis adopts Oracle Fusion Data Intelligence

dunnhumby is a British data science company founded in 1989, renowned for its pioneering work in customer data analysis and loyalty strategies. The firm’s meteoric rise to fame began in the mid-1990s when it partnered with Tesco, one of the UK’s most prominent supermarkets, helping to launch the Tesco Clubcard. dunnhumby’s innovative approach enabled retailers to better understand shopper behavior, tailor marketing efforts, and transform the retail landscape across the UK and beyond.

By leveraging data-driven insights, dunnhumby quickly became a household name in the world of consumer analytics, earning a reputation for turning raw data into actionable intelligence for leading brands and retailers globally.

Making business digital

Since 2016, dunnhumby has used Oracle Fusion Cloud Applications to digitize back-office functions in Finance and HR, in addition to ServiceNow and most recently Salesforce. They needed to go beyond the operational reporting to provide finance, HR, and sales deeper insights into revenue, pipeline, retention, and costs. They chose to implement Oracle Fusion Data Intelligence (FDI), recognizing the value of being able to analyze curated Fusion Applications data across multiple domains—ERP and HR—and blend it with other external data.

Evolving analytics at dunnhumby

Giorgio Ticinelli has more than 20 years of experience providing analytics on customer data for retailers.

He joined dunnhumby in 2012 and in 2019 became BI Leader in Corporate Systems. He established the company’s first specialist BI team, working in partnership with the finance, sales, and HR teams. Prior to that, business leaders relied on transactional reporting and many Excel spreadsheets and pivot tables created by external contractors for business insight. Sales was using a Microsoft Access solution that required manual work every day. After a period using on-premises Oracle BI Applications and Oracle Business Intelligence Enterprise Edition (OBIEE), Ticinelli and the team switched to cloud solutions with Oracle Analytics Cloud (OAC) and Autonomous Data Warehouse. Ticinelli noted, “The experience we gained with Oracle on-premises BI Applications was not wasted, because it was easy to transition the skills we had and the requirements we’d gathered from OBIEE to OAC.” When FDI was launched in 2020, providing cloud-based prebuilt content for Oracle Fusion Cloud Applications, dunnhumby switched.

Meeting the analytic needs of the business took collaboration

dunnhumby wanted to enhance governance around how Fusion Applications data was used for analytics. “Instead of contractors reacting to every new request for information, we were a central team through which requests from users for new analysis could be channeled,” said Ticinelli. “We strive to listen to every request, so people are not frustrated; they know we try our best to deliver.” In addition to creating a central team to handle requests, Ticinelli’s group drove strong collaboration with business stakeholders to establish requirements and priorities for additional dashboards beyond the packaged FDI content. Early on, champions were identified in Finance and HR, who were critically important in guiding development and becoming FDI experts themselves. They also took full advantage of the self-service capabilities in FDI to create their own content, delving into new data sources as they became available. They ultimately became advocates who encouraged their peers, guiding them how to take advantage of FDI.

Often, it’s important to focus first on the users’ experience using a dashboard, to deliver visually what they need. In this case, the team takes a “dashboard first” approach. Once they have a dashboard design the way users want it, they figure out what data, what joins, what the schema should look like, and they construct the query for the dashboard.

To get users up-to-speed and getting value, the team created a training path and posted how-to videos and PDFs on new capabilities as they were released; they also made use of Oracle Learning & Development to deliver it to users. In addition, they created a Microsoft Teams channel dedicated to BI.

Getting an accurate view of the company and its employees

One of the most important dashboards for HR was the Position dashboard. It provides an up-to-date view of every position in the company. Ticinelli explained, “Every people manager can review many people they have, how many are leaving, how many are going to be hired, how many are still vacant.  HR used to do this in Excel, and it was complex for them to keep current and accurate, especially for big teams with over a thousand positions. Now it’s easy, using FDI and the HR content direct from Fusion HCM.”

Motivating staff

The team created the “Well Dunn” dashboard, combining HR data with data from the employee recognition system. This recognition system enables managers to award perks to people for demonstrating excellence in a variety of values important to dunnhumby. Now the leadership can see the distribution of perks, by value for each team, country, or organisation.

Another example is from Learning and Development, joining training data with HR to be able to analyze what training has been completed, by whom, and which are the most popular. There are also a Diversity and Inclusion dashboard and a dashboard for recruitment KPI’s.

Connecting the dots in Fusion data

For Ticinelli, the data pipelines from Fusion Applications represent the biggest value of FDI. “In my experience, BI is not just about creating dashboards; it’s about making sure that the data is correct. If the data is correct, then building dashboards using OAC is pretty easy. The Fusion Applications data model is one of the most complex database designs I’ve ever seen; building a data warehouse on top from scratch would be difficult. With FDI, you can leverage Oracle know-how. You don’t have to worry about how to join dozens of tables. You can take any subject area, let’s say the HCM workforce or the procurement subject area, and easily start to build visualizations without much effort. If we need to blend data by joining two subject areas, that’s also easy with FDI. And it’s fast—performance is excellent.”

Blending Fusion Applications data with Salesforce and ServiceNow

The pipelines to the Salesforce schema prebuilt by Oracle enable dunnhumby to blend Salesforce and Fusion Applications data to measure forecasted revenue and opportunities in FDI. The company has also built pipelines directly to FDI from ServiceNow, loading data daily. At present this capability is with the Business Optimization team, but soon, anybody managing a team will be able to easily review their service tickets and see satisfaction scores. Then dunnhumby will have an analytics hub: one URL that decision makers can use to get all the information they need from finance, HR, Salesforce and ServiceNow.

Business intelligence AI at dunnhumby

There’s a thirst in the user community to be able to augment dashboards with conversational interfaces, enabling them to ask any question of their data on the fly. Ticinelli and his team have worked on that, collaborating with Oracle to help improve the FDI AI Assistant. In the area of Learning & Development, they have used Python to engineer transformations of unstructured data so they can ingest it into FDI to improve decision making.

In the future, Ticinelli plans to make more use of the new Oracle AI Data platform to improve processes for structured and unstructured data transformation, and to provide a single unified platform to manage AI initiatives.

What does it mean for dunnhumby to GROW?

As a data science organization, dunnhumby’s commitment to data and analytics is foundational and a big part of their company culture. They capture their guiding principles using the acronym GROW:

G — Growth (adoption + scope): BI adoption is increasing significantly across functions.

R — Return (ROI + efficiency): Estimated annual ROI with payback in few years, plus automation saving recurring manual work.

O — Outcomes (business impact): Supported decision making for HR, L&D, Sales and Finance, and provided actionable information for teams that would have been blind without data.

W — Ways of working (people + governance + power users): Success factors explicitly include collaboration, leadership alignment, clear objectives, and power users helping spread adoption. With support from senior management, adoption of FDI at dunnhumby has been rapid, up 25% year-on-year. Now around 600 managers, decision makers, and business analysts in Finance, HR and Sales rely on FDI daily to provide the insight needed to make the best data-driven decisions.

Learn More

Learn more about dunnhumby’s data, analytics, and AI work and see their winning data visualizations in the Oracle Analytics & AI Community.