“From Spreadsheets to Data Warehouse – Transforming A Division’s Data Approach”
|Media, Agency, Sports & Entertainment||Publicis Sports & Entertainment||Broadcast, Media, Audience Data, Data Cleansing||2022||Data Discovery, Data Pipelines, Data Migration, Reporting, Data Visualisation|
Publicis Groupe is one of the largest media and advertising agencies in the world, with 80k employees and $13bn in global revenues. Within Publicis Groupe, the Sports & Entertainment division (“PSE”) offers brands and rights holders unique insights into broadcast media, audience segmentation and match/game analysis.
PSE is responsible for data aggregation and reporting for some of the world’s largest brands, rights holders and match organizations including UEFA, FIFA, Eurosport and the International Olympic Committee, to name a few.
As the primary data processing entity, they receive sports events viewership audience data from 500+ TV broadcasters and streaming platforms globally, all unstructured and in their own localized unique forms and formats.
Traditionally, PSE could not rely on broadcasters to supply data in the right format – usually these are “slices” or manual extracts which are time consuming and rarely in a standardized template. It was the job of teams of analysts to manually triage flat files (spreadsheets), manually merge them and prepare presentations.
PSE required a new approach that used modern cloud data infrastructure, automation, ETL processes and data warehousing capabilities. They selected Bomisco to create the necessary infrastructure to receive unstructured flat files, apply highly complex media industry rules, transform / cleanse the files, detect anomalies and run a continuous data delivery pipeline into multiple data warehouses (including BigQuery and Snowflake) and data visualization tools (PowerBI and Tableau).
Bomisco combined a DataOps approach with modern cloud data tools to deliver uplift to PSE by creating the “Data Desk”. As files came in, they would be automatically triaged, flagged and ingested into the platform. Over the last 12 months, Bomisco has developed a unique global and source-based ruleset that transforms the data at massive scale, automates error detection and delivers significantly cleaner outputs for QA.
The cleansed data is then pushed into a data warehouse which can then be leveraged by clients and other data insights tools such as PowerBI for interactive reporting and data visualization. This reduced the time-to-market, upgraded their data capabilities and enabled an entirely new DataOps capability.
Bomisco was able to save thousands of hours of manual data handling by providing a managed data pipeline. Publicis relies on Bomisco to deliver uplift and work towards their goal of becoming a best-in-class data team that combines the power of big data with deep media and sports experience to its clients.