Jo Louter, Consumer Data & Analytics Senior Manager
Rahul Kishore, Sr Manager, Industrialization and Automation Services
Ryan Odonovan, Data Architect
Jan Wielebinski, Senior Data Engineer
Piotr Kosewski, Data Engineer
Neil Abragimowicz, Senior Data Engineer
Enes Bolfidan, Data Mapper
Michael Brooks, Audiences & Performance Marketing Senior Manager
Ceren Akillioglu, Global Audience Analyst
Veerdhwaj Rathore, Data Scientist & Asst Manager, Industrialization and Automation Services
Shreyas Shelke, Data Scientist
Country: United Kingdom
Unilever is one of the world’s leading suppliers of Beauty & Personal Care, Home Care, and Foods & Refreshment products, with sales in over 190 countries and products used by 3.4 billion people every day. We have 148,000 employees and generated sales of €52.4 billion in 2021. Over half of our footprint is in developing and emerging markets. We have around 400 brands found in homes all over the world – including iconic global brands like Dove, Lifebuoy, Knorr, Magnum, OMO, and Surf; and other brands such as Love Beauty & Planet, Hourglass, Seventh Generation, and The Vegetarian Butcher.
Unilever’s ambition is to build personalized digital relationships with its 3.4 billion consumers across the globe. By doing so, we can tailor our communications to relevant consumers at the right time and through the right channel, ultimately improving our return on investment.
To accomplish this, we needed to enable our 38 Digital Hubs to understand the behaviors, attributes, interests, and demographics of each individual consumer that engages with our digital channels. This would allow us to create consumer audiences that could be executed across our digital media landscape.
Our challenges in doing this were primarily:
The initial phase was to overhaul the data completely. Using a lean team of four data engineers and a data architect, we created a set of primary data pipelines in Dataiku to produce a semantic layer of our consumer data. This allowed us to efficiently ensure that huge amounts of data are cleansed, standardized, logically structured, and accurate. Using Dataiku’s scenarios functionalities, these pipelines are automated and scheduled. Resulting in no human intervention being required to ensure that the data is robust and available 24/7/365. We leveraged Dataiku’s ML capability to cleanse this data.
Our next focus was to join our disparate sets of consumer data to provide an enriched Unified View of Consumers (UVC). Our ad tech provides us with deterministic matches of consumer identifiers across different digital consumer touchpoints. And by using Dataiku, we could easily stitch all the consumer’s descriptive data to these matched identifiers to create a deeper understanding of each consumer. Dataiku enables a seamless transfer of this data into our enterprise-wide visualization tool.
Finally, utilizing Dataiku’s webapp capability, we built a simple, point-and-click interface (Audience Creation Tool, aka ACT) for non-technical users, to create audiences based on our UVC and a consumer’s unique set of behaviors, attributes, interests, and demographics. ACT transfers the audience to the relevant execution channel via a single click, powered by customized API connectors built in Dataiku.
Business area enhanced: Communication/Strategy/Competitive Intelligence/IT/Cybersecurity/Data/Marketing/Sales/Customer Relationship Management
Use case stage: In Production
The value of our semantic layer, UVC & ACT has been realized in 3 key areas:
At every step, Dataiku has been instrumental in achieving our aims. Its flexibility and capabilities have been the reason we were able to overcome the challenges we faced along the way.
Dataiku’s breadth of functionality has been a key factor in our success, enabling us to easily service a wide variety of user profiles, from non-technical to experienced data scientists and data engineers, in one tool and with one source of data. With all these functionalities in one central tool, Dataiku has not only improved collaboration across our teams but has also improved our efficiency in leveraging our consumer data. It has done this by enabling us to centralize the end-to-end processes, from experimentation and exploration to optimizing for scale, productionizing, and maintenance.
In addition, this centralization enabled us to embed our data privacy and ethics principles into their design, ensuring that our users can seamlessly use this data in the appropriate manner.
Value range: Tens of millions $ and above